Journal of Media Psychology 2020 by Hogrefe

Volume 32 / Number 1 / 2020

Journal of

Media Psychology

Journal of Media Psychology

Editor-in-Chief Christoph Klimmt Associate Editors Nick D. Bowman Allison Eden Jesse Fox Özen Odağ Catalina Toma Ivar Vermeulen

Theories, Methods, and Applications

Hogrefe OpenMind Open Access Publishing? It’s Your Choice! Your Road to Open Access Authors of papers accepted for publication in any Hogrefe journal can now choose to have their paper published as an open access article as part of the Hogrefe OpenMind program. This means that anyone, anywhere in the world will – without charge – be able to read, search, link, send, and use the article for noncommercial purposes, in accordance with the internationally recognized Creative Commons licensing standards.

The Choice Is Yours 1. O pen Access Publication: The final “version of record” of the article is published online with full open access. It is freely available online to anyone in electronic form. (It will also be published in the print version of the journal.) 2. Traditional Publishing Model: Your article is published in the traditional manner, available worldwide to journal subscribers online and in print and to anyone by “pay per view.” Whichever you choose, your article will be peer-reviewed, professionally produced, and published both in print and in electronic versions of the journal. Every article will be given a DOI and registered with CrossRef.

www.hogrefe.com

How Does Hogrefe’s Open Access Program Work? After submission to the journal, your article will undergo exactly the same steps, no matter which publishing option you choose: peer-review, copy-editing, typesetting, data preparation, online reference linking, printing, hosting, and archiving. In the traditional publishing model, the publication process (including all the services that ensure the scientific and formal quality of your paper) is financed via subscriptions to the journal. Open access publication, by contrast, is financed by means of a one-time article fee (€ 2,500 or US $3,000) payable by you the author, or by your research institute or funding body. Once the article has been accepted for publication, it’s your choice – open access publication or the traditional model. We have an open mind!

Journal of

Media Psychology Theories, Methods, and Applications

Volume 32/ Number 1/2020

Editor-in-Chief Editorial Assistants

Associate Editors

Editorial Board

Publisher

Production Subscriptions Advertising/Inserts ISSN Copyright Information

Publication Subscription Prices

Payment

Electronic Full Text Abstracting/Indexing Services

Journal of Media Psychology (2020), 32(1)

Christoph Klimmt, Department of Journalism and Communication Research, Hanover University of Music, Drama and Media, 30539 Hanover, Germany (E-mail christoph.klimmt@ijk.hmtm-hannover.de) Sophie Bruns, Katharina Emde-Lachmund, Katharina Knop-Hülß, and Sabine Reich, Department of Journalism and Communication Research, Hanover University of Music, Drama and Media, 30539 Hanover, Germany (E-mail jmp@ijk.hmtm-hannover.de) Nick D. Bowman, Texas Tech University, Lubbock, TX, USA, E-mail nick.bowman@ttu.edu Allison Eden, Michigan State University, East Lansing, MI, USA, E-mail edenalli@msu.edu Jesse Fox, The Ohio State University, Columbus, OH, USA, E-mail fox775@osu.edu Özen Odaǧ, Touro College, Berlin, Germany, E-mail oezen.odag@touroberlin.de Catalina Toma, University of Wisconsin-Madison, Madison, WI, USA, E-mail ctoma@wisc.edu Ivar Vermeulen, Vrije Universiteit Amsterdam, The Netherlands, E-mail i.e.vermeulen@vu.nl Eun-Ju Lee (Seoul, South Korea) Markus Appel (Würzburg, Germany) Jörg Matthes (Vienna, Austria) Florian Arendt (München, Germany) Peter Nauroth (Marburg, Germany) Omotayo Banjo (Cincinnati, OH, USA) German Neubaum (Duisburg-Essen, Germany) Anne Bartsch (München, Germany) Anne Oeldorf-Hirsch (Mansﬁeld, CT, USA) Gary Bente (Cologne, Germany) Jochen Peter (Amsterdam, The Netherlands) Paul Bolls (Columbia, MO, USA) Daniel Pietschmann (Chemnitz, Germany) Johannes Breuer (Cologne, Germany) Robert F. Potter (Bloomington, IN, USA) Jennings Bryant (Tuscaloosa, AL, USA) Arthur A. Raney (Tallahassee, USA) Caleb Carr (Normal, IL, USA) Leonard Reinecke (Mainz, Germany) Elizabeth Cohen (Morgantown, WV, USA) Diana Rieger (Munich, Germany) Enny Das (Nijmegen, The Netherlands) Meghan Sanders (Baton Rouge, LA, USA) Kevin Durkin (Glasgow, UK) Frank Schneider (Mannheim, Germany) Nicole Ellison (Ann Arbor, MI, USA) Frank Schwab (Würzburg, Germany) Malte Elson (Bochum, Germany) Stephan Schwan (Tübingen, Germany) David Ewoldsen (Columbus, OH, USA) Michael Slater (Columbus, OH, USA) Christopher Ferguson (DeLand, FL, USA) Kaveri Subrahmanyam (Los Angeles, CA, USA) Jesse Fox (Columbus, OH, USA) Ron Tamborini (East Lansing, MI, USA) Sabine Glock (Wuppertal, Germany) Sabine Trepte (Hohenheim, Germany) Melanie Green (Chapel Hill, NC, USA) Mina Tsay-Vogel (Boston, MA, USA) Matthew Grizzard (Buffalo, NY, USA) Dagmar Unz (Würzburg-Schweinfurt, Germany) Dorothée Hefner (Hannover, Germany) Sonja Utz (Tübingen, Germany) Shirley S. Ho (Singapore) Sebastián Valenzuela (Santiago de Chile, Chile) Matthias Hofer (Zurich, Switzerland) Brandon van der Heide (East Lansing, MI, USA) Juan José Igartua (Salamanca, Spain) Christian von Sikorski (Vienna, Austria) Jimmy Ivory (Blacksburg, VA, USA) Peter Vorderer (Mannheim, Germany) Jeroen Jansz (Rotterdam, The Netherlands) Patrick Weber (Hohenheim, Germany) Julia Kneer (Rotterdam, The Netherlands) René Weber (Santa Barbara, CA, USA) Elly Konijn (Amsterdam, The Netherlands) Stephan Winter (Duisburg-Essen, Germany) Maja Krakowiak (Colorado Springs, CO, USA) Mike Yao (Hong Kong, ROC) Nicole Krämer (Duisburg-Essen, Germany) Annie Lang (Bloomington, IN, USA) Hogrefe Publishing, Merkelstr. 3, 37085 Göttingen, Germany, Tel. +49 551 99950-0, Fax +49 551 99950-425, E-mail publishing@hogrefe.com, Web www.hogrefe.com North America: Hogrefe Publishing, 361 Newbury Street, 5th Floor, Boston, MA 02115, USA, Tel. (857) 880-2002, E-mail publishing@hogrefe.com Regina Pinks-Freybott, Hogrefe Publishing, Merkelstr. 3, 37085 Göttingen, Germany, Tel. +49 551 99950-0, Fax +49 551 99950-425, E-mail production@hogrefe.com Hogrefe Publishing, Herbert-Quandt-Str. 4, 37081 Göttingen, Germany, Tel. +49 551 99950-999, Fax +49 551 99950-998, E-mail zeitschriftenvertrieb@hogrefe.de Melanie Beck, Hogrefe Publishing, Merkelstr. 3, 37085 Göttingen, Germany, Tel. +49 551 99950-423, Fax +49 551 99950-425, E-mail marketing@hogrefe.com ISSN-L 1864-1105, ISSN-Print 1864-1105, ISSN-Online 2151-2388 Ó 2020 Hogrefe Publishing. This journal as well as the individual contributions and illustrations contained within it are protected under international copyright law. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, microﬁlming, recording or otherwise, without prior written permission from the publisher. All rights, including translation rights, reserved. Published in 4 issues per annual volume. The Journal of Media Psychology is the continuation of Zeitschrift für Medienpsychologie (ISSN 1617-6383), the last annual volume of which (Volume 19) was published in 2007. Calendar year subscriptions only. Rates for 2020: Institutions – from US $353.00 /€ 272.00 (print only; pricing for online access can be found in the journals catalog at hgf.io/journalscatalog); Individuals – US $195.00 / € 144.00 (print & online; all plus US $16.00 / € 12.00 postage & handling). Single copies – US $93.00 / € 73.00 (plus postage & handling) Payment may be made by check, international money order, or credit card, to Hogrefe Publishing, Merkelstr. 3, 37085 Göttingen, Germany. US and Canadian subscriptions can also be ordered from Hogrefe Publishing, 361 Newbury Street, 5th Floor, Boston, MA 02115, USA. The full text of Journal of Media Psychology is available online at http://econtent.hogrefe.com and in PsycARTICLES. Abstracted/indexed in Current Contents/Social and Behavioral Sciences (CC/S&BS), Social Sciences Citation Index (SSCI), IBR, IBZ, PsycINFO, PsycLit, PSYNDEX, and Scopus. 2018 Impact Factor 1.514, 5-year Impact Factor 1.648, Journal Citation Reports (Clarivate Analytics, 2019)

Ó 2020 Hogrefe Publishing

Contents Editorial

At the Start of 2020 – Transitions in the Editorial Team

Christoph Klimmt Original Articles

How Perceived Processing Fluency Inﬂuences the Illusion of Knowing in Learning From TV Reports Fabian A. Ryffel and Werner Wirth

Spoilers Go Bump in the Night: Impacts of Minor and Major Reveals on Horror Film Enjoyment Benjamin K. Johnson, Angel Udvardi, Allison Eden, and Judith E. Rosenbaum

Erratum

Correction to Johnson, Udvardi, Eden, & Rosenbaum, 2019

Original Article

Research Report

Exploring the Effects of Personal Information in Television News: A Cognitive Approach Miguel Barreda-Ángeles, Julia Virgós, Maddalena Fedele, and Alexandre Pereda-Baños

Call for Papers

‘‘Theoretical Consolidation and Innovation in Media Psychology’’: A Special Issue of the Journal of Media Psychology, Curated by the Editors Nicholas D. Bowman, Allison Eden, Jesse Fox, Christoph Klimmt, Ózen Oda g, Ivar Vermeulen, and Catalina Toma

Ó 2020 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1)

Editorial At the Start of 2020 – Transitions in the Editorial Team Christoph Klimmt Department of Journalism and Communication Research, Hanover University of Music, Drama and Media, Germany

“It is the same with people as it is with riding a bike. Only when moving can one comfortably maintain one’s balance.” Einstein to his son, Feb 5, 1930 (With permission: The Hebrew University of Jerusalem) The new year has once again begun with profound changes of the team that is curating the Journal of Media Psychology (JMP). First, the good news: Allison Eden has started her term as new Associate Editor in January. She is Associate Professor at Michigan State University in East Lansing (USA) and brings rich experience with theory and research in media psychology to her new role. Allison knows JMP well, as she is holding a long-standing record of authorship and reviewer service. So we are excited to welcome another top media psychologist of international renown in the JMP team! Even more good news: Özen Odağ has joined the circle of Associate Editors of JMP in February. She is Associate Professor at Touro College in Berlin (Germany) and has also been working with the journal extensively in the past, both as author and reviewer. Özen will enrich JMP’s portfolio of expertise, particularly because of her excellent and unique track record in cultural and gender dimensions of media psychology. We are delighted that Özen has agreed to serve as Associate Editor and are confident to handle the high number of manuscripts that JMP is receiving with the expanded editorial team. Unfortunately, these new beginnings come also with a good-bye. After a decade of service to the journal – acting as an Editorial Assistant from 2010 to 2014, and since then as Associate Editor – Diana Rieger (LMU Munich, Germany) has departed towards new academic challenges. Diana has worked with Editors-in-Chief Gary Bente, Nicole Krämer, and me. She has curated more than 90 manuscripts as Associate Editor and contributed actively to the development of the journal on multiple dimensions. She has energized the journal team with kindness, humor, and strong engagement. But most importantly, we celebrate Diana’s philosophy of

Ó 2020 Hogrefe Publishing

editing a journal, because it can serve as a blueprint of what kind of academic organization JMP wants to be for the global media psychology community: Diana has always worked through the submissions that were assigned to her with outstanding dilligence, curiosity, and the interest to identify their strengths and potentials. She has developed countless ideas and suggestions to support authors in refining their argumentation, and, maybe most importantly, she has practiced a benevolent, kind, and fair mode of interaction with authors, reviewers, and Hogrefe Publishing alike. This is what all JMP editors are striving for: Excellence that firmly connects to benevolence. Hopefully, Diana’s editorial service will thus be a lighthouse example for future JMP editors – it certainly is for the current team. We deeply appreciate her sustained service, welcome her among the members of the Editorial Board (which to join she has fortunately agreed to), and wish her all the best for the future.

Published online February 25, 2020 Christoph Klimmt Department of Journalism and Communication Research Hanover University of Music, Drama and Media Expo Plaza 12 30539 Hanover Germany christoph.klimmt@ijk.hmtm-hannover.de

Christoph Klimmt (PhD, 2004) studied media management at the Department of Journalism and Communication Research (IJK) of Hannover University. From 2007 to 2010, he served as an assistant professor in the Department of Communication, University of Mainz. Since 2010, he has been a professor of communication science at IJK Hannover. His research interests include media effects and processes, entertainment, and digital games.

Journal of Media Psychology (2020), 32(1), 1 https://doi.org/10.1027/1864-1105/a000275

Original Article

How Perceived Processing Fluency Influences the Illusion of Knowing in Learning From TV Reports Fabian A. Ryffel and Werner Wirth Department of Communication and Media Research, University of Zurich, Switzerland Abstract: The present two-study work aims to contribute to an understanding of the causes and consequences of perceived processing fluency. Regarding its causes, the experimental studies (N1 = 399; N2 = 337) found that features of television reports (e.g., footage used, background music) can heighten perceptions of processing fluency. Regarding its consequences, it was found that heightened perceived fluency biases metacognitive judgments. Specifically, considering perceived knowledge in relation to actual knowledge, recipients experiencing fluency have been found to overestimate their knowledge about the issue depicted in the experimental stimuli. The resulting illusion of knowing was particularly pronounced under conditions of low involvement, indicating that the so-called ease-of-processing heuristic underlies the effect. Keywords: processing fluency, perceived knowledge, knowledge acquisition, illusion of knowing

Individuals experience information processing differently. Sometimes processing information feels easy and effortless; in some cases, information processing feels more difficult. In the following, we argue that this sense of perceived ease or difficulty associated with information processing is critical for actual as well as perceived knowledge. Specifically, we assume that perceived ease associated with information processing influences both actual and perceived knowledge, to a different extent, however. As a consequence, a gap between perceived and actual knowledge can arise leading individuals to overestimate their knowledge about the issue in question. Thus, “media exposure might not increase actual knowledge as much as it might enhance the impression of one’s own knowledge [. . .]” (Park, 2001, p. 419). Perceptions of ease associated with information processing are captured in the concept of perceived processing fluency (e.g., Alter & Oppenheimer, 2008; Oppenheimer, 2008). Perceived processing fluency can take place at various levels: “All cognitive activity, from lower order processes such as perception to higher order processes such as deductive reasoning, can be placed along a continuum from fluent to disfluent” (Oppenheimer & Alter, 2013, p. 85). Accordingly, perceptual fluency or lower-order processing fluency has been conceptualized as the ease of identifying surface features of stimuli. Perceptual fluency occurs, for example, when the physical form of a stimulus is identified. This type of fluency has been shown to be influenced by parameters such as clarity, figure-to-ground Journal of Media Psychology (2020), 32(1), 2–13 https://doi.org/10.1027/1864-1105/a000250

contrast, or simplicity (e.g., Reber, Schwarz, & Winkielman, 2004). Conceptual fluency or higher-order processing fluency reflects the ease of mental operations going on when the semantic meaning of a stimulus is identified and cognitively related to extant knowledge structures (e.g., Winkielman, Schwarz, Fazendeiro, & Reber, 2003). Conceptual fluency is influenced by variables such as predictability or context congruity (e.g., McGlone & Tofighbakhsh, 2000). Perceived processing fluency influences various judgments (Unkelbach & Greifeneder, 2013). Among others, perceived fluency has been found to exert an influence on metacognitive judgments. Specifically, perceived processing fluency can lead humans to overestimate their knowledge about the issue in question (e.g., Koriat, Bjork, Sheffer, & Bar, 2004; Kornell, Rhodes, Castel, & Tauber, 2011). This socalled illusion of knowing, in turn, might lead some people to publicly express their opinion (e.g., Neuwirth, 2000), even though they are rather uninformed about the topic they comment on. Additionally, people who come to the erroneous conclusion that they know enough about an issue might also avoid relevant media coverage (e.g., Kaye & Johnson, 2002). Thus, illusions of knowing might also be important for subsequent media selection. Therefore, perceived fluency is of great importance to media psychology. Nevertheless, perceived fluency has not been examined in the context of media reception. In previous research, perceptions of processing fluency have been experimentally manipulated using several stimulus manipulations (for an Ó 2018 Hogrefe Publishing

F. A. Ryffel & W. Wirth, The Illusion of Knowing

overview, see Alter & Oppenheimer, 2009). For instance, perceived fluency has been varied by manipulating font type (e.g., Alter, Oppenheimer, Epley, & Eyre, 2007), contrast (e.g., Reber, Winkielman, & Schwarz, 1998), difficulty of pronunciation (e.g., Alter & Oppenheimer, 2006), or linguistic difficulty (e.g., Oppenheimer, 2006). However, these manipulations were implemented in rather simple stimuli. But since they can be generated by “nearly any form of cognitive processing” (Oppenheimer, 2008, p. 238), we suggest that fluency perceptions might also occur when more complex stimuli such as television reports are processed. Similar to stimuli used in previous research, television reports can be enriched with features facilitating processing. We expect that visual (footage, editing techniques) and acoustic (background music, narrator voice) packaging will – if used in the proper extent – contribute to fluency experiences during television reception (Wirth, 2013). To begin with, the footage used might actuate fluency perceptions. Specifically, a clear image language in accordance with the storyline of a report should enhance perceived fluency. In other words, visual congruity is expected to increase fluency. Moreover, editing techniques such as slow motion or color effects are expected to influence fluency. When these stylistic means are deployed in accordance with the storyline and support the understanding of a report’s content, they are expected to heighten perceived fluency. Likewise, the acoustic packaging of television reports might support fluency experiences. Specifically, acoustic congruity should increase perceived processing fluency: When dramatic background music is used during dramatic scenes this might facilitate following the story plot as compared with the same scene with incongruous or without background music (Cohen, 2015). Similarly, a narrator voice matching the storyline in terms of tonality should increase fluency. These assumptions are in line with research showing that divergences between text and pictures in television reports reduce their comprehensibility (e.g., Brosius, Donsbach, & Birk, 1996). However, when the use of the described means is exaggerated, perceived processing fluency is expected to decrease. That is, when the image language will literally underscore every aspect of the spoken word, this might create a cognitive overload impeding fluency experiences. Likewise, excessive acoustic packaging is expected to decrease processing fluency. Empirical support for this claim can be drawn from research showing that stimuli containing a greater number of elements are more complex (e.g., Kreitler, Zigler, & Kreitler, 1974) and that media environments overstraining recipients decrease perceived ease of use (Liu, Liao, & Pratt, 2009). Hence, we expect the relation between visual and acoustic packaging and perceived processing fluency to be curvilinear altogether. When the discussed means are Ó 2018 Hogrefe Publishing

used in the right proportion, they jointly contribute to fluency experiences during television reception. However, we assume that fluency perceptions occur not only at one certain point on the manipulation continuum, but rather on a broader range. Not until the use of the audiovisual features is exaggerated is perceived fluency expected to decrease. Hypothesis 1 (H1): The relation between audiovisual packaging and perceived processing fluency is curvilinear. Fluency perceptions will be highest when the audiovisual features are used moderately and will be relatively low when they are absent or their use is exaggerated.

Fluency Perceptions and Illusions of Knowing In the previous section, we argued that perceived processing fluency during media reception entail illusions of knowing. But how do illusions of knowing arise? The illusion of knowing has been defined as “the gap between self-perceived and actual knowledge” (Park, 2001, p. 421) about a particular issue. That is, two parties contribute to illusions of knowing: perceived and actual knowledge. We argue that both of these concepts are influenced by fluency perceptions, although to a different extent.

Fluency Perceptions and Perceived Knowledge Previous research found that processing fluency influences various judgments (for a review, see Unkelbach & Greifeneder, 2013). For instance, fluency perceptions have been shown to influence judgments of truth (e.g., Unkelbach, 2007), liking (e.g., Forster, Leder, & Ansorge, 2013), or aesthetics (e.g., Reber et al., 2004). Likewise, perceived fluency has been found to influence – or rather bias – metacognitive judgments. Specifically, perceived fluency has been shown to affect perceived knowledge (e.g., Koriat et al., 2004). This finding has been explained through the so-called ease-of-processing heuristic. Specifically, it has been argued that metacognitive judgments are based on immediate subjective impressions rather than on theorybased, in-depth analyses of metacognitive processes. In such appraisal processes, the ease-of-processing heuristic is a piece of information that is easily available and therefore affects metacognitive judgments (Kornell et al., 2011). In other words, the ease-of-processing heuristic is considered the application of fluency perceptions when Journal of Media Psychology (2020), 32(1), 2–13

F. A. Ryffel & W. Wirth, The Illusion of Knowing

making judgments.1 But why would fluency affect judgments of learning in a positive way? This is due to the fact that fluency is a nonspecific experience and its interpretation is learned (Unkelbach, 2006). In some situations, learning is associated with fluency and these situations might suggest that processing fluency always indicates learning gains. Consider, for instance, a student solving math problems. After a few trials she or he might comprehend how to solve a certain type of problem. Comprehension in turn will lead the student to experience fluency when she or he encounters another problem of that type. Hence, in this case, fluency indeed indicates learning and attributing heightened fluency to successful learning would lead the student to appropriately estimate her or his learning gain. When fluency is, however, triggered by sources other than comprehension (e.g., features of television reports), it can be misattributed to learning (Unkelbach, 2006). In other words, human beings notice that they experience fluency when they have acquired knowledge and learn the corresponding interpretation of fluency: Fluency experiences mark learning gains. After having learned that knowledge acquisition is accompanied by fluency, people wrongly conclude that fluency will always indicate learning gains – because that is the only interpretation of the nonspecific experience they have learned. That is, people are unaware of the fact that fluency can be triggered by sources such as stimulus features and will not invariably indicate learning gains. Therefore, fluency misattributions disappear when unrelated fluency sources are identified (e.g., Schwarz et al., 1991). Hypothesis 2 (H2): Perceived processing fluency during television reception positively affects perceived knowledge. As described earlier, perceived fluency serves as a heuristic in metacognitive judgments (Kornell et al., 2011). The weight people place on such heuristics is particularly high under conditions of low involvement. Specifically, when task involvement is low, people monitor their metacognitive processes less accurately and rely on heuristics when making metacognitive judgments (e.g., Chaiken, Liberman, & Eagly, 1989). Accordingly, fluency misattributions should be particularly high when recipients process information under conditions of low task involvement. Hypothesis 3 (H3): The influence of perceived processing fluency on perceived knowledge will be stronger when involvement is low as compared with high involvement.

Fluency Perceptions and Knowledge Acquisition Drawing on the limited capacity model of motivated mediated message processing (LC4MP; Lang, 2000), we expect that fluent processing of mediated messages will be accompanied by heightened knowledge acquisition. In the LC4MP it is assumed that recipients process information making use of a limited and fixed capacity of cognitive resources. Processing of mediated information involves three subprocesses: encoding, storage, and retrieval. Encoding has been conceptualized as the process of “selecting information from the environment and creating a mental representation of that information” (Lang, 2009, p. 194) and includes operations such as perception and comprehension. The subsequent process, storage, consists of creating links between newly encoded information and extant knowledge structures in order to store it in memory. Finally, retrieval is the subprocess where previously stored information is activated in order to understand or contextualize incoming information. The processes operate simultaneously and each subprocess requires some capacity. Moreover, the human cognitive system only has finite capacity and the resources available need to be divided between the subprocesses (Lang, 2000). Therefore, when messages are encoded effortlessly – as is the case when messages are processed fluently – only few resources are allocated to message encoding. Hence, messages that are encoded fluently will leave more cognitive capacity for storage of the incoming information, which will occur along with enhanced knowledge acquisition. Hypothesis 4 (H4): Perceived processing fluency during television reception occurs along with heightened actual knowledge acquisition.

Differences Between Actual and Perceived Knowledge Combining the lines of argumentation, we expect the gap between actual and perceived knowledge to be particularly large when perceived fluency is high. Specifically, when processing is not facilitated, actual knowledge acquisition is expected to be relatively low. But since fluency – and consequently the ease-of-processing heuristic – would be absent in this case, recipients should estimate their knowledge gains accurately to be relatively low. Thus, the gap between actual and perceived knowledge should be small

Some authors refer to processing fluency as a cue rather than a heuristic (e.g., Unkelbach & Greifeneder, 2013). However, the influence of fluency on judgments is similar, regardless of the specific terminology.

Journal of Media Psychology (2020), 32(1), 2–13

Ó 2018 Hogrefe Publishing

F. A. Ryffel & W. Wirth, The Illusion of Knowing

Hypothesis 5 (H5): Perceived processing fluency during television reception positively affects the gap between perceived and actual knowledge (i.e., the illusion of knowing). Since individuals place greater weight on heuristics under conditions of low involvement, the effect should be particularly pronounced when involvement is low. Hypothesis 6 (H6): The influence of perceived processing fluency on the gap between perceived and actual knowledge (i.e., the illusion of knowing) will be stronger when involvement is low as compared with high involvement. In sum, we do not expect fluency and consequently illusions of knowing to occur when no features promoting processing fluency are included in television reports. When the features are used moderately, we expect fluency and the resulting illusion of knowing to be particularly high. When the use of the audiovisual features is exaggerated, however, fluency and the resulting illusion of knowing should decrease. These assumptions guiding our hypotheses are depicted in Figure 1.

Study 1 Method To test Hypotheses 1–3, Study 1 employed a 4 2 (audiovisual features [absent, moderate, prominent, exaggerated] involvement [low, high]) factorial design. The study was conducted online and participants were randomly assigned to the experimental conditions. Since we were unable to control whether participants used search engines to find the correct answers, we did not employ an objective knowledge measure. Thus, Study 1 focused on the causes of fluency perceptions and their effect on perceived knowledge. To make sure that the stimulus videos were displayed in adequate quality, participation was not possible using smartphones or tablet computers, but only using home computers or laptops. To assure that participants could hear Ó 2018 Hogrefe Publishing

Extent of Actual/Perceived Knowledge

since both meet at relatively low levels. Conversely, experiencing fluency will occur along with heightened knowledge acquisition. But if fluency is experienced, recipients should make inordinately high estimations of their own knowledge gains. That is, even though fluency should be accompanied by actual knowledge acquisition, recipients experiencing fluency overestimate their knowledge gains owing to the ease-of-processing heuristic. Hence, higher estimations of one’s own knowledge gains as a result of perceived fluency would only be partially correct.

Illusion of Knowing

Actual Knowledge Perceived Knowledge

Audiovisual Packaging Figure 1. Expected relation between audiovisual packaging, actual knowledge, and perceived knowledge. This relation is expected to be mediated by perceived processing fluency and moderated by involvement during message processing.

the sound of the stimulus videos, a listening test was conducted at the beginning of the experiment.

Participants The sample of Study 1 consisted of 399 adults from the German-speaking part of Switzerland, aged between 18 and 79 (Mage = 45.03, SDage = 15.12; 64.2% females).

Material Television reports specifically created for the experiment were used as experimental stimuli. The reports dealt with a controversy on lead hunting ammunition. Specifically, the reports described that the majority of hunters in Switzerland used lead ammunition even though ammunition made from other materials is available. The reports stressed that the release of lead residues is unavoidable when lead ammunition is used. Regarding potential negative consequences, several arguments were presented. For instance, the reports stressed that lead residues are dangerous for scavengers because lead has been found to accumulate in their bodies. Moreover, the reports explained that the lead residues might pose a risk to people eating venison. On the opposite side, it was brought forward, for instance, that ammunition made from alternative materials is much more expensive than lead ammunition. Moreover, the reports stressed that ammunition made from other materials would not kill the hunted animals effectively leading them to suffer slow and painful deaths. The four stimulus versions were entirely realized by professionals. The raw material used for the reports was shot by a camera team of the Swiss broadcasting company Journal of Media Psychology (2020), 32(1), 2–13

(SRF). The footage was originally used for the science show Einstein. To create four different versions of the report, the raw material shot by the camera team was made available to the research team. After combing through the material, scripts for the four stimulus versions were written by the research team (see Manipulation of Independent Variables). The stimuli were then realized by a professional film editor. In addition, a professional narrator, who recorded the voice-over in a recording studio, was hired. Thus, the stimuli used for the experiment met the highest standards of quality and completely looked like real television reports.

Manipulation of Independent Variables In the four stimulus versions, audiovisual features were used to a varying degree (absent, moderate, prominent, exaggerated). Additionally, participants were randomly assigned to one of two involvement conditions (low, high). Audiovisual Packaging Four versions of the same television report were created. Regarding their content, the versions were identical. However, regarding audiovisual packaging, the versions varied considerably. This was achieved by systematically varying the footage used, editing techniques, background music, and the tonality of the voice-over narration. The first stimulus version (absent) mainly consisted of very long takes of landscapes and woods where hunters go shooting. The report did not show hunting scenes and did not include any footage showing the different types of ammunition. Moreover, only a few quotes of hunters and experts were included in the first version, the largest share of information was conveyed by the narrator. Additionally, in the first version neither visual effects nor background music was used and the narrator provided information in a monotonous style. In the second stimulus version (moderate), the footage used was closer to the storyline. That is, for instance, when the properties of the different types of ammunition were explained, the particular type of ammunition was shown in the picture. Similarly, the report contained some music that was congruent to the storyline. For example, when the narrator explained that alternative ammunitions could make the hunted animals die slowly and painfully, sad music faded in. In addition, to underscore the controversy, proponents (i.e., hunters) and opponents (i.e., experts such as biologists) presented some of the arguments regarding lead ammunition in their direct speech. The narrator voice used in the second stimulus was more lively and congruent with the information conveyed. The third stimulus version (prominent) underscored most of the information presented with congruent footage and music. For instance, when the report depicted that lead Journal of Media Psychology (2020), 32(1), 2–13

F. A. Ryffel & W. Wirth, The Illusion of Knowing

ammunition killed very effectively, a shooting scene from the first-person perspective was shown. The scene featured thrilling background music, sound effects (i.e., heart beats), and crosshairs were layered over the deer that was about to be shot. Additionally, to clearly underscore the controversy, the scenes where proponents and opponents of lead ammunition presented their arguments were of contrasting color grading. The narrator voice in Version 3 was highly emotional, but remained congruent with the storyline at all times. In the fourth stimulus version (exaggerated), literally any aspect of the report was underscored by congruent footage, background music, and sound effects. Exaggerating the use of audiovisual features resulted in an overly fast cut. The main problem in creating four stimulus versions was to conceive four equidistant levels of audiovisual packaging. To achieve this objective, the stimuli were developed in three steps. First, scripts for the four stimulus versions were written and repeatedly pretested in written form. Second, draft versions of the stimulus reports were created by the research team. Again, these versions were pretested and repeatedly edited in response to the results. After the objective of equidistance was achieved for the draft versions, the professional film editor created the stimuli used for the experiment. These versions were also pretested and slightly modified in response to the pretest results. Involvement To manipulate involvement, participants read an instruction prior to watching the stimulus video. In the high involvement condition, participants were informed that the report they were about to see was of utmost importance. They were instructed to pay close attention to the facts conveyed, since they would have to perform a knowledge test after stimulus exposure. In the low involvement condition, participants were informed that the report they were about to see might not be of interest to them. They were instructed not to pay too much attention to the details conveyed.

Measures Audiovisual Packaging To create a treatment check for audiovisual packaging, participants were asked to what extent the report they had seen contained the manipulated features; specifically, a semantic differential for each manipulated stimulus feature (footage, editing techniques, background music, and narrator voice; α = .79). Involvement Involvement was measured using three items (e.g., “I concentrated on the report’s content while watching it”) on Ó 2018 Hogrefe Publishing

F. A. Ryffel & W. Wirth, The Illusion of Knowing

5-point Likert scales ranging from 1 = totally disagree to 5 = totally agree (α = .79). Perceived Processing Fluency The perceived processing fluency measurement consisted of two semantic differentials successfully applied in previous research (Lee & Aaker, 2004; Ryffel & Wirth, 2016). Participants responded to the prompt, “The information presented was...” on a 5-point Likert-type scale from 1 = easy to understand to 5 = difficult to understand, as well as on a 5-point Likert-type scale from 1 = easy to process to 5 = difficult to process (ρ = .83). Perceived Knowledge Perceived knowledge was measured using nine items (e.g., “I am aware of the topic’s details”) on a 5-point Likert scale ranging from 1 = totally disagree to 5 = totally agree (α = .84; Mattheiß et al., 2013).

Results Manipulation Check To test whether the manipulation of audiovisual packaging was successful, the corresponding treatment check index was submitted to a two-way ANOVA, using both manipulations as independent variables. Results show that perceptions of audiovisual packaging increased linearly between the stimulus versions. Specifically, participants who watched the first version (absent; M = 2.06, SD = 0.70) reported that they had seen and heard less of the manipulated features than did participants who were presented with the second version (moderate; M = 2.64, SD = 0.62). Similarly, differences in the expected direction were obtained between Version 2 and 3 (prominent; M = 3.10, SD = 0.51), and 3 and 4 (exaggerated; M = 3.53, SD = 0.64), resulting in a highly significant effect of audiovisual packaging, F(3, 391) = 103.013, p < .001, ηp2 = .44. Involvement, F(1, 391) = 0.712, ns, and the interaction between audiovisual packaging and involvement, F(3, 391) = 1.45, ns, did not have an effect. Additionally, it was analyzed whether the versions differed from one another by calculating post hoc tests using Bonferroni correction. The post hoc tests revealed that Version 1 differed from Version 2, Mdiff = –0.59, 95% CI = 0.76– 0.41, p < .001, Version 2 differed from Version 3, Mdiff = 0.46, 95% CI = 0.63– 0.29, p < .001, and Version 3 differed from Version 4, Mdiff = 0.43, 95% CI = 0.60– 0.25, p < .001. Given the fact that the mean differences obtained only varied between 0.43 and 0.59, the versions can be 2

considered equidistant and the increase can be considered linear. Hence, the manipulation was successful.2 To test whether the involvement manipulation was successful, the involvement index was submitted to a twoway ANOVA using the same independent variables as before. Results revealed that participants in the low involvement condition (M = 3.64, SD = 0.81) were less involved than participants in the high involvement condition (M = 3.91, SD = 0.72, F(1, 391) = 12.34, p < .001, ηp2 = .03. Audiovisual packaging did not have an influence on involvement during stimulus exposure, F(3, 391) = 0.78, ns. Additionally, no interaction between the two independent variables was obtained, F(3, 391) = 1.174, ns. Hence, the involvement manipulation produced a small effect in the intended direction.

Test of Hypotheses To test Hypotheses 1–3, a regression model with audiovisual packaging as the independent variable, perceived fluency as a mediator, the experimental factor involvement as a moderator, and perceived knowledge as the dependent variable (see Figure 2) was run using the PROCESS macro by Hayes (2013; Model 15, bootstrap estimation with 1,000 samples, bias-corrected). To model the hypothesized curvilinear effect of audiovisual packaging on perceived fluency, the independent variable was dummy-coded (absent = 0, moderate = 1, prominent = 1, exaggerated = 0). Setting both the moderate and prominent version to 1 and contrasting them with the absent and exaggerated versions reflects the idea that fluency perceptions will occur on a broader range on the audiovisual packaging continuum. In other words, we expect heightened fluency perceptions to arise in the moderate as well as in the prominent version. This model revealed a highly significant curvilinear effect of audiovisual packaging on perceived fluency (b = 0.25, SE = .08, p = .001), supporting Hypothesis 1. The same holds for Hypothesis 2: Fluency perceptions were found to have an effect on perceived knowledge (b = 0.94, SE = .15, p < .001). Finally, an interaction between involvement and fluency was obtained (b = 0.30, SE = .09, p = .001). The conditional effects indicated that the effect of fluency on perceived knowledge was stronger in the low involvement condition (b = .16, SE = .05, 95% CI = 0.07–0.28) than in the high involvement condition (b = .08, SE = .03, 95% CI = 0.04–0.16), thus supporting Hypothesis 3. Moreover, the model revealed an effect of involvement on perceived knowledge (b = 1.09, SE = .38, p = .004). Including the mediator variable diminished the direct effect to nonsignificance (b = 0.20, SE = .22, ns; total

To test whether audiovisual packaging unintentionally influenced participants’ interest in the reports, we conducted confound checks. The confound checks for both studies clearly indicated that participants interest did not differ between versions.

Ó 2018 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1), 2–13

F. A. Ryffel & W. Wirth, The Illusion of Knowing

Table 1. Means (SD) of perceived processing fluency and perceived knowledge by experimental condition Stimulus version Variable

Absent

Moderate

Prominent

Exaggerated

Low involvement

4.13 (0.83)

4.42 (0.64)

4.30 (0.60)

4.14 (0.76)

High involvement

3.94 (0.96)

4.19 (0.71)

4.15 (0.71)

3.95 (0.77)

Low involvement

3.83 (0.79)

4.13 (0.70)

4.16 (0.65)

4.12 (0.64)

High involvement

3.80 (0.84)

3.73 (0.77)

3.53 (0.77)

3.71 (0.97)

Perceived processing fluency

Perceived knowledge

effect: b = 0.43, SE = .13). The descriptive statistics for Study 1 are depicted in Table 1.

Discussion Hypotheses 1–3 were supported in Study 1. Thus, Study 1 was the first study to empirically demonstrate that perceived fluency can occur when more complex stimuli such as television reports are processed. Moreover, Study 1 demonstrated that perceived fluency acts as a heuristic for metacognitive judgments. This conclusion is underpinned by the finding that the influence of perceived fluency on perceived knowledge is particularly strong under conditions of low involvement: Participants made more use of the heuristic when involvement was low as compared with high involvement. In Study 2 we investigated whether heightened perceptions of one’s own knowledge as a consequence of fluency perceptions are valid. Specifically, in Study 2 perceived knowledge was considered in relation to the recipients’ actual knowledge. Thus, all of the six hypotheses presented earlier were tested in Study 2.

Study 2 Method Study 2 employed the same 4 2 (audiovisual features [absent, moderate, prominent, exaggerated] involvement [low, high]) factorial design applied in Study 1. But since actual knowledge was measured in Study 2, it had to be conducted as a laboratory experiment in order to make sure that participants did not use search engines to find the correct answers. Participants took part in single sessions and were randomly assigned to the experimental conditions.

Participants The sample of Study 2 consisted of 337 adults from the German-speaking part of Switzerland, aged between 18 and 78 (Mage = 29.60, SDage = 13.49; 67.4% females). Journal of Media Psychology (2020), 32(1), 2–13

Material Again, television reports were created by professionals for the purpose of the experiment. The reports for Study 2 dealt with a controversy on livestock-guarding dogs: The stimuli described that more and more shepherds keep livestockguarding dogs to protect their flock against the growing number of wolves living in Switzerland. Moreover, it was discussed that the guarding dogs pose a danger to hikers crossing the flock’s territory, because the dogs protect the sheep against all creatures that do not belong to the flock. The reports featured experts who explained the dogs’ nature and gave recommendations on how to behave when crossing flocks of sheep in the mountains. Again, the raw material used for the reports in Study 2 was shot by a camera team of the Swiss broadcasting company and originally used for the science show Einstein.

Manipulation of Independent Variables The manipulations of the independent variables were identical to those employed in Study 1. That is, on the one hand, audiovisual features were used to a varying degree (absent, moderate, prominent, exaggerated). However, regarding their content, the four versions were identical. On the other hand, the involvement manipulation (low, high) was performed by presenting participants with the same instruction used in Study 1.

Measures The measures already employed in Study 1 were used in Study 2. Specifically, the measures for audiovisual packaging (α = .83), involvement (α = .72), perceived processing fluency (ρ = .79), and perceived knowledge (α = .80) were included in the questionnaire. Additionally, an objective knowledge measure was created and the illusion of knowing was operationalized. Objective Knowledge After the perceived knowledge measurement, participants’ actual knowledge about facts conveyed in the report was Ó 2018 Hogrefe Publishing

F. A. Ryffel & W. Wirth, The Illusion of Knowing

measured using four closed- and four open-ended questions. The facts requested were related to specific content of the report (e.g., “What was the name of the alp the report was shot on?”). To answer the closed-ended questions, participants had to choose from four options. To make sure that participants did not choose the correct answer by chance, a fifth option labeled “I don’t know” was included. Additionally, participants were instructed not to guess when answering the questions. Participants received one point for correct answers and zero points for incorrect answers. The answers to the open-ended questions were content-analytically evaluated by the research team. Correct answers were coded as 1, answers that were almost correct were coded as 0.66, somewhat correct answers were coded as 0.33, and completely incorrect answers were coded as 0. An intercoder reliability test based on 50 randomly selected answers was conducted and produced satisfactory reliability (Cohen’s r = 0.87). To create an objective knowledge index, the points participants received were summed up. Illusion of Knowing To create the illusion of knowing score, a difference index between perceived and objective knowledge was created. Since the objective and perceived knowledge measures differed in range, they had to be normalized prior to creating the index. Hence, both variables were normalized using a feature scaling method. That is, both variables were rescaled to only take on values between 0 and 1. The normalized objective knowledge scores were then subtracted from the normalized perceived knowledge scores. Hence, positive values on the resulting index indicate overestimation, while negative values indicate underestimation of one’s own knowledge.

Results Manipulation Check To test whether audiovisual packaging worked as intended, the treatment check index was submitted to a two-way ANOVA using both manipulations as independent variables. Results suggest that the manipulation was successful: Participants who were assigned the first stimulus version (absent; M = 2.33, SD = 0.78) scored lower on the treatment check index than did participants who watched the second version (moderate; M = 2.81, SD = 0.71). Participants in the third stimulus condition (prominent; M = 3.34, SD = 0.69) scored lower on the according index than did participants who watched version four (exaggerated; M = 4.07, SD = 0.66). Hence, a highly significant effect of audiovisual packaging was obtained, F(3, 329) = 87.79, p < .001, ηp2 =

Ó 2018 Hogrefe Publishing

.45. Moreover, neither involvement, F(1, 329) = 3.28, ns, nor the interaction between audiovisual packaging and involvement, F(3, 329) = 0.656, ns, reached significance. Post hoc tests with Bonferroni correction revealed that Version 1 differed from Version 2, Mdiff = 0.49, 95% CI = 0.70– 0.27, p < .001, Version 2 differed from Version 3, Mdiff = 0.53, 95% CI = 0.74– 0.33, p < .001, and Version 3 differed from Version 4, Mdiff = 0.72, 95% CI = 0.95– 0.51, p < .001. Again, the mean differences only varied inconsiderably, ranging from 0.49 to 0.72. Hence, the increase between the versions is considered linear and the versions are considered equidistant. To test the involvement manipulation, the involvement index was submitted to the same two-way ANOVA. Results show that participants in the low involvement condition (M = 3.73, SD = 0.68) were less involved than participants in the high involvement condition (M = 4.06, SD = 0.68), F(1, 329) = 20.33, p < .001, ηp2 = .05. Audiovisual packaging did not have an influence on involvement during stimulus exposure, F(3, 329) = 1.34, ns. Additionally, no interaction between the independent variables was revealed, F(3, 329) = 0.79, ns. Hence, the manipulation of involvement produced a small effect in the intended direction.

Test of Hypotheses To test Hypotheses 1–3 again, the same regression model applied in Study 1 (see Figure 2) was run using the PROCESS macro by Hayes (2013; Model 15, bootstrap estimation with 1,000 samples, bias-corrected). Again, the analysis revealed a curvilinear effect of audiovisual packaging on perceived processing fluency (b = 0.35, SE = .08, p < .001). Moreover, fluency had an effect on perceived knowledge (b = 0.51, SE = .07, p < .001), and the interaction between involvement and perceived processing fluency occurred as predicted in Hypothesis 3 (b = 0.25, SE = .10, p = .012). The conditional effects indicated that the effect of fluency on perceived knowledge was stronger when involvement was low (b = .18, SE = .05, 95% CI = 0.10–0.29) than when involvement was high (b = .09, SE = .04, 95% CI = 0.03–0.19). Moreover, involvement was found to influence perceived knowledge, (b = 1.00, SE = .42, p = .019). Again, the direct effect of audiovisual packaging on perceived knowledge was insignificant after including the mediator in the model (b = 0.08, SE = .11, ns; total effect: b = 0.26, SE = .09). To test the relationship between perceived processing fluency and actual knowledge, the same model with participants’ actual knowledge scores as dependent variable was rerun (see Figure 3). As predicted in Hypothesis 4, this analysis revealed that when perceived processing fluency was high, actual

Journal of Media Psychology (2020), 32(1), 2–13

F. A. Ryffel & W. Wirth, The Illusion of Knowing

Perceived Processing Fluency

Stimulus Version

–0.30** (SE = .09)

0.35** (SE = .08)

0.94** (SE = .15)

Perceived Knowledge

0.20 (SE = .22) (0.43 [SE = .13])

(dummy-coded)

Figure 2. Moderated mediation model (Hayes, 2013) testing Hypotheses 1–3 (**p < .01).

Perceived Processing Fluency

0.35** (SE = .08)

Stimulus Version (dummy-coded)

Involvement –0.06 (SE = .03)

0.06** (SE = .02)

Actual Knowledge

0.07 (SE = .04) (0.09 [SE = .03])

Figure 3. Moderated mediation model (Hayes, 2013) testing Hypothesis 4 (and 1; **p < .01).

knowledge was also heightened (b = 0.06, SE = .02, p = .009). Additionally, involvement was found to positively influence actual knowledge (b = 0.35, SE = .14, p = .014). The interaction between fluency and involvement (b = 0.06, SE = .03, ns) and the direct effect (b = 0.07, SE = .04, ns; total effect: b = 0.09, SE = .03) were insignificant. To test Hypotheses 5 and 6 the same model was rerun using the illusion of knowing index as a dependent variable (see Figure 4). This model shows that perceptions of fluency during processing of television reports promote illusions of knowing. Specifically, the model revealed that perceived processing fluency affects the illusion of knowing, (b = 0.46, SE = .07, p < .001). This finding is in line with Hypothesis 5.

Stimulus Version

Involvement –0.22* (SE = .10)

0.46** (SE = .07)

Illusion of Knowing

0.04 (SE = .12) (0.20 [SE = .09])

(dummy-coded)

Figure 4. Moderated mediation model (Hayes, 2013) on the illusion of knowing, testing Hypotheses 5 and 6 (and 1; *p < .05, **p < .01).

Normalized Values of Actual/Perceived Knowledge

0.25** (SE = .08)

Perceived Processing Fluency

Involvement

0.74

Actual Knowledge Perceived Knowledge

0.72 0.7

Illusion of Knowing

0.68 0.66 0.64 0.62 0.6 1

Stimulus Version

Figure 5. Empirical relation between audiovisual packaging, actual knowledge, and perceived knowledge. This relation has been found to be mediated by perceived processing fluency and moderated by involvement during message processing.

Additionally, this influence is stronger when involvement is low (b = .16, SE = .04, 95% CI = 0.09–0.26) as compared with high involvement (b = .08, SE = .04, 95% CI = 0.02–0.18), resulting in a significant interaction effect, (b = 0.22, SE = .10, p = .036). This result supports Hypothesis 6. Again, no direct effect occurred (b = 0.04, SE = .12, ns; total effect: b = 0.20, SE = .09). To visualize the illusion of knowing obtained in Study 2, the normalized values of perceived and actual knowledge are depicted in Figure 5; the descriptive statistics for Study 2 are depicted in Table 2.

Table 2. Means (SD) of perceived processing fluency, perceived knowledge, and objective knowledge by experimental condition Stimulus version Variable

Absent

Moderate

Prominent

Exaggerated

Low involvement

3.87 (1.07)

4.44 (0.62)

4.52 (0.62)

4.18 (0.79)

High involvement

4.19 (0.79)

4.44 (0.67)

4.42 (0.52)

4.13 (0.89)

Low involvement

3.56 (0.70)

3.75 (0.75)

3.95 (0.63)

3.53 (0.89)

High involvement

3.53 (0.89)

3.94 (0.73)

3.78 (0.78)

3.37 (0.81)

Low involvement

4.28 (1.90)

4.66 (1.49)

5.25 (1.59)

4.99 (1.58)

High involvement

5.32 (1.99)

5.42 (1.80)

5.51 (1.85)

5.19 (1.66)

Perceived processing fluency

Perceived knowledge

Objective knowledge

Journal of Media Psychology (2020), 32(1), 2–13

Ó 2018 Hogrefe Publishing

F. A. Ryffel & W. Wirth, The Illusion of Knowing

Discussion In addition to findings from Study 1, Study 2 demonstrated that fluency is related to actual knowledge. Moreover, to investigate whether heightened perceptions of knowledge as a consequence of perceived processing fluency are valid, actual knowledge was considered in relation to recipients’ perceived knowledge. The corresponding analysis revealed that fluency perceptions bias the recipients’ perceptions of their knowledge, promoting an illusion of knowing. This illusion was particularly pronounced under conditions of low involvement. This finding indicates that perceived processing fluency biases metacognitive judgments as a heuristic.

General Discussion On the one hand, the present paper contributes to knowledge about the causes of processing fluency. In contrast to previous studies, perceptions of fluency were induced using relatively complex stimuli. Specifically, it was shown that fluency can be heightened by specific features of television reports. However, future research should look more closely at the specific features of mediated messages promoting processing fluency. For instance, it is possible that visual features contribute to perceptions of ease to a greater extent than do acoustic features or vice versa. Additionally, the particular visual or acoustic features might contribute to fluency experiences to a varying degree. To harness the concept of processing fluency for media effects research, its causes need to be understood in more detail. On the other hand, we examined the consequences of fluency perceptions and found that fluency entails illusions of knowing. These illusions become visible in our data at the transition from Stimulus Version 2 to 3. Fluency already increases at moderate levels of audiovisual packaging. Therefore, as a consequence of the ease-of-processing heuristic, perceived knowledge also goes up for the second stimulus version. Actual knowledge acquisition, however, does not considerably increase at moderate levels of audiovisual packaging. Actual knowledge does not increase until Stimulus Version 3. The resulting gap is a clear display of the illusion of knowing. These findings are of utmost relevance because estimations of one’s own knowledge guide media selection (e.g., Kaye & Johnson, 2002). Specifically, it can be assumed that politically active citizens strive to have good knowledge of current topics. These recipients might come to wrongly conclude that they know enough about an issue as a result of fluency. If that is the case, they might avoid media coverage on that issue because they unjustly believe not to need any

Ó 2018 Hogrefe Publishing

more knowledge on the issue in question. Conversely, one might argue that fluency as a pleasant state might be a source of attentional inertia (Hawkins et al., 2002). That is, experiencing fluency might lead some people to stick with television content they can process fluently. Moreover, since television is particularly effective in creating fluency perceptions, it is also likely to create illusions of knowing. As a consequence, viewers might stick to television as a primary source of political information because it permanently gives them the feeling of being well-informed. Future research should investigate the consequences of fluency and subsequent metacognitive judgments for media selection. Additionally, the illusion of knowing might also lead some people to speak out publicly about certain topics, even though they are not able to contribute to the debate (e.g., Neuwirth, 2000). If we imagine discussions on social networking sites, for instance, some people might behave brashly as a consequence of their illusion of knowing, which might discourage other discussants. As a result, some public debates might be taken over by rather uninformed people. Nevertheless, given that perceived fluency also promoted actual knowledge acquisition in our data, some scholars might find it desirable to provide media content facilitating processing. Future research should further investigate the potential consequences of illusions of knowing. The present results should be interpreted within the studies’ potential limitations. First, it should be noted that fluency was induced by manipulating television reports. This raises the question of whether the present results can be generalized to all kinds of media (e.g., newspaper articles). Future research should investigate whether the effects of the present study also occur for other types of media. Second, in the present paper, illusions of knowing were observed for an issue the vast majority of participants were unfamiliar with. Hence, one might conclude that illusions of knowing only occur at low levels of knowledge. Future research should investigate whether the effects also occur for more prominent issues. Third, future research should further investigate the consequences of perceived processing fluency during media reception. In the present paper, we found that fluency during media reception influences metacognitive judgments. Based on these results, it seems probable that fluency during media reception also influences other kinds of judgments. For instance, looking at the effects of media-induced fluency on judgments of truth seems particularly promising (Unkelbach, 2007). Research Transparency Statement The authors are willing to share their data, analytics methods, and study materials with other researchers. The material is available upon request.

Journal of Media Psychology (2020), 32(1), 2–13

References Alter, A. L., & Oppenheimer, D. M. (2006). Predicting short-term stock fluctuations by using processing fluency. Proceedings of the National Academy of Sciences, 103, 9369–9372. https://doi. org/10.1073/pnas.0601071103 Alter, A. L., & Oppenheimer, D. M. (2008). Easy on the mind, easy on the wallet: The roles of familiarity and processing fluency in valuation judgments. Psychonomic Bulletin & Review, 15, 985– 990. https://doi.org/10.3758/PBR.15.5.985 Alter, A. L., & Oppenheimer, D. M. (2009). Uniting the tribes of fluency to form a metacognitive nation. Personality and Social Psychology Review, 13, 219–235. https://doi.org/10.1177/ 1088868309341564 Alter, A. L., Oppenheimer, D. M., Epley, N., & Eyre, R. N. (2007). Overcoming intuition: Metacognitive difficulty activates analytic reasoning. Journal of Experimental Psychology: General, 136, 569–576. https://doi.org/10.1037/0096-3445.136.4.569 Brosius, H. B., Donsbach, W., & Birk, M. (1996). How do text-picture relations affect the informational effectiveness of television newscasts? Journal of Broadcasting & Electronic Media, 40, 180–195. https://doi.org/10.1080/08838159609364343 Chaiken, S., Liberman, A., & Eagly, A. H. (1989). Heuristic and systematic information processing within and beyond the persuasion context. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 212–252). New York, NY: The Guilford Press. Cohen, A. J. (2015). Congruence-association model and experiments in film music: Toward interdisciplinary collaborations. Music and the Moving Image, 8, 5–24. https://doi.org/ 10.5406/musimoviimag.8.2.0005 Forster, M., Leder, H., & Ansorge, U. (2013). It felt fluent, and I liked it: Subjective feeling of fluency rather than objective fluency determines liking. Emotion, 13, 280–289. https://doi. org/10.1037/a0030115 Hawkins, R. P., Pingree, S., Bush Hitchon, J., Gilligan, E., Kahlor, L., Gorham, B. W., & Serlin, R. C. (2002). What holds attention to television? Strategic inertia of looks at content boundaries. Communication Research, 29, 3–30. https://doi.org/10.1177/ 0093650202029001001 Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach: Methodology in the social sciences. New York, NY: The Guilford Press. Kaye, B. K., & Johnson, T. J. (2002). Online and in the know: Uses and gratifications of the Web for political information. Journal of Broadcasting & Electronic Media, 46, 54–71. https://doi.org/ 10.1207/s15506878jobem4601_4 Koriat, A., Bjork, R. A., Sheffer, L., & Bar, S. K. (2004). Predicting one’s own forgetting: The role of experience-based and theory-based processes. Journal of Experimental Psychology: General, 133, 643–656. https://doi.org/10.1037/0096-3445. 133.4.643 Kornell, N., Rhodes, M. G., Castel, A. D., & Tauber, S. K. (2011). The ease-of-processing heuristic and the stability bias.Dissociation memory, memory beliefs, and memory judgments. Psychological Science, 22, 787–794. https://doi.org/10.1177/ 0956797611407929 Kreitler, S., Zigler, E., & Kreitler, H. (1974). The complexity of complexity. Human Development, 17, 54–73. https://doi.org/ 10.1159/000271333 Lang, A. (2000). The limited capacity model of mediated message processing. Journal of Communication, 50, 46–70. https://doi. org/10.1111/j.1460-2466.2000.tb02833.x

Journal of Media Psychology (2020), 32(1), 2–13

F. A. Ryffel & W. Wirth, The Illusion of Knowing

Lang, A. (2009). The limited capacity model of motivated mediated message processing. In R. L. Nabi & M. B. Oliver (Eds.), The SAGE handbook of media processes and effects (pp. 193–204). Thousand Oaks, CA: SAGE. Lee, A. Y., & Aaker, J. L. (2004). Bringing the frame into focus: The influence of regulatory fit on processing fluency and persuasion. Journal of Personality and Social Psychology, 86, 205–218. Liu, S. H., Liao, H. L., & Pratt, J. A. (2009). Impact of media richness and flow on e-learning technology acceptance. Computers and Education, 52, 599–607. https://doi.org/10.1016/ j.compedu.2008.11.002 Mattheiß, T., Weinmann, C., Löb, C., Rauhe, K., Bartsch, K., Roth, F. S., & Vorderer, P. (2013). Political learning through entertainment – only an illusion? How motivations for watching TV political talk shows influence viewers’ experiences. Journal of Media Psychology, 25, 171–179. https://doi.org/10.1027/18641105/a000100 McGlone, M. S., & Tofighbakhsh, J. (2000). Birds of a feather flock conjointly (?): Rhyme as reason in aphorisms. Psychological Science, 11(5), 424–428. https://doi.org/10.1111/1467-9280. 00282 Neuwirth, K. (2000). Testing the spiral of silence model: The case of Mexico. International Journal of Public Opinion Research, 12, 138–159. https://doi.org/10.1093/ijpor/12.2.138 Park, C. Y. (2001). News media exposure and self-perceived knowledge: The illusion of knowing. International Journal of Public Opinion Research, 13, 419–425. https://doi.org/10.1093/ ijpor/13.4.419 Oppenheimer, D. M. (2006). Consequences of erudite vernacular utilized irrespective of necessity: Problems with using long words needlessly. Applied Cognitive Psychology, 20, 139–156. https://doi.org/10.1002/acp.1178 Oppenheimer, D. M. (2008). The secret life of fluency. Trends in Cognitive Sciences, 12(6), 237–241. https://doi.org/10.1016/j. tics.2008.02.014 Oppenheimer, D. M., & Alter, A. L. (2013). Disfluency sleeper effect: Disfluency today promotes fluency tomorrow. In C. Unkelbach & R. Greifeneder (Eds.), The experience of thinking. How the fluency of mental processes influences cognition and behavior (pp. 85–97). New York, NY: Psychology Press. Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is beauty in the perceiver’s processing experience? Personality and Social Psychology Review, 8, 364–382. https://doi.org/10.1207/ s15327957pspr0804_3 Reber, R., Winkielman, P., & Schwarz, N. (1998). Effects of perceptual fluency on affective judgments. Psychological Science, 9, 45–48. https://doi.org/10.1111/1467-9280.00008 Ryffel, F. A., & Wirth, W. (2016). Heart versus mind: How affective and cognitive message frames change attitudes. Social Psychology, 47, 52–62. https://doi.org/10.1027/18649335/ a000257 Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauerschatka, H., & Simons, A. (1991). Ease of retrieval as information: Another look at the availability heuristic. Journal of Personality and Social Psychology, 61, 195–202. https://doi.org/10.1037/ 0022-3514.61.2.195 Unkelbach, C. (2006). The learned interpretation of cognitive fluency. Psychological Science, 17, 339–345. https://doi.org/ 10.1111/j.1467-9280.2006.01708.x Unkelbach, C. (2007). Reversing the truth effect: Learning the interpretation of processing fluency in judgments of truth. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33, 219–230. https://doi.org/10.1037/0278-7393. 33.1.219

Ó 2018 Hogrefe Publishing

F. A. Ryffel & W. Wirth, The Illusion of Knowing

Unkelbach, C., & Greifeneder, R. (2013). A general model of fluency effects in judgment and decision making. In C. Unkelbach & R. Greifeneder (Eds.), The experience of thinking. How the fluency of mental processes influences cognition and behavior (pp. 11– 32). New York, NY: Psychology Press. Winkielman, P., Schwarz, N., Fazendeiro, T., & Reber, R. (2003). The hedonic marking of processing fluency: Implications for evaluative judgment. In J. Musch & K. C. Klauer (Eds.), The psychology of evaluation: Affective processes in cognition and emotion (pp. 189–217). Mahwah, NJ: Erlbaum. Wirth, W. (2013). Zur Rezeption von Infotainment: Informationsund Unterhaltungserleben als Konstituenten der InfotainmentRezeption [On the reception of infotainment: Information and entertainment experiences as constituents of infotainment reception]. In M. Dohle & G. Vowe (Eds.), Politische Unterhaltung–Unterhaltende Politik: Forschung zu Medieninhalten, Medienrezeption und Medienwirkungen (pp. 57–93). Cologne, Germany: Halem.

Fabian A. Ryffel is a postdoctoral research and teaching associate of the division Media Psychology & Effects at the Department of Communication and Media Research (IKMZ) of the University of Zurich, Switzerland.

Werner Wirth is a full professor for empirical communication and media research and head of the division Media Psychology & Effects at the Department of Communication and Media Research (IKMZ) of the University of Zurich, Switzerland.

History Received February 23, 2017 Revision received July 8, 2018 Accepted July 10, 2018 Published online December 19, 2018 Werner Wirth Department of Communication and Media Research University of Zurich 8050 Zurich Switzerland w.wirth@ikmz.uzh.ch

Ó 2018 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1), 2–13

Original Article

Spoilers Go Bump in the Night Impacts of Minor and Major Reveals on Horror Film Enjoyment Benjamin K. Johnson1 , Angel Udvardi1, Allison Eden2, and Judith E. Rosenbaum3 1

Department of Advertising, University of Florida, Gainesville, FL, USA

Department of Communication, Michigan State University, East Lansing, MI, USA

Department of Communication and Journalism, University of Maine, Orono, ME, USA Abstract: Spoilers are frequently a source of concern for entertainment audiences. Online discussions, promotional materials, and reviews can all potentially reveal pertinent information about story plotlines, presumably ruining suspense and enjoyment. Despite these common apprehensions, recent experimental evidence suggests that narrative spoilers have modest and inconsistent effects on enjoyment and other audience responses. In this study, we investigate the implications of spoilers for horror films, a genre reliant on suspenseful thrills. Each participant was exposed to multiple scenes from a single horror film, and spoilers were manipulated at different stages of the narrative for both minor and major plot points. Results indicate no main effects of spoilers for enjoyment, transportation, suspense, processing fluency, or reactance. However, need for affect moderated positive enjoyment effects of spoilers for smaller plot points such as scary moments or sudden twists. These results indicate that those who value the thrills of horror films may receive enjoyment from the anticipation produced by minor spoilers about smaller plot points. Keywords: spoilers, enjoyment, processing fluency, narrative, horror film

Is a frightening film still scary if you know the ending? The question of whether spoilers ruin enjoyment is still under investigation, with extant research producing conflicting results. On the one hand, the commonly held assumption that uncertainty is key to suspense and enjoyment (Zillmann, Hay, & Bryant, 1975), and that spoilers thus ruin narratives (Hassoun, 2013), is supported by recent research showing harmful albeit modest effects of spoilers on the enjoyment of short stories (Johnson & Rosenbaum, 2015; Levine, Betzner, & Autry, 2016). On the other hand, studies have also demonstrated that spoilers can in fact improve enjoyment by facilitating processing fluency (Ellithorpe & Brookes, 2018; Leavitt & Christenfeld, 2011, 2013). The present study extends existing work on spoilers by investigating a genre that is known for its reliance on suspense and fear to create enjoyment: horror. In addition, this study examines the role played by the placement of the spoiler in the story and by the spoiler’s magnitude in people’s experience of a horror movie. The current line of research into the impact of spoilers on user experience started with the experimental finding that spoilers, instead of ruining, actually enhance people’s enjoyment of short stories (Leavitt & Christenfeld, 2011). This single study was followed by a flurry of replications that yielded a series of, at times, conflicting findings. Some studies show that spoilers improved enjoyment (e.g., Ellithorpe Journal of Media Psychology (2020), 32(1), 14–25 https://doi.org/10.1027/1864-1105/a000252

& Brookes, 2018; Leavitt & Christenfeld, 2013), others show harm to enjoyment (e.g., Daniel & Katz, 2018; Johnson & Rosenbaum, 2015; Levine et al., 2016), and others suggest people tend to overestimate the negative impact of spoilers on their enjoyment (Yan & Tsang, 2016). Further work illustrated that processing fluency mediated effects of spoilers on audience evaluations (Leavitt & Christenfeld, 2013) and that the traits of need for cognition (NFC) and need for affect (NFA) had small moderating effects on spoiler acceptance and spoiler enjoyment, respectively (Rosenbaum & Johnson, 2016). Not all spoilers are created equal, either. Topolinksi (2014) found that when there was a shorter delay in the time between presentation of a spoiler and exposure to the narrative, the more the spoiler would decrease enjoyment. This is at odds with Levine et al. (2016), who demonstrated that spoilers presented in advance of exposure harmed enjoyment, while spoilers presented during the narrative did not influence enjoyment at all. When it comes to the nature of the spoiler, Yan and Tsang (2016) showed that, contrary to viewer expectations, spoilers that were focused on the outcome of a story did not impact enjoyment, whereas spoilers that focused on the plot decreased enjoyment. Additionally, identifying a spoiler as such as well as the subtlety or bluntness of a spoiler both have small effects on narrative enjoyment (Johnson & Rosenbaum, 2018; Ó 2019 Hogrefe Publishing

B. K. Johnson et al., Minor and Major Horror Spoilers

Levine et al., 2016). The present study investigates the influence of spoilers on audience enjoyment, accounting for the location (beginning, middle, or end) as well as the magnitude (big or small) of the spoiled plot point. In addition, we test mediating roles of suspense, reactance, and processing fluency, as well as moderating roles of the NFC, NFA, and horror fandom.

Horror, Spoilers, and Responses While previous studies have considered the role of genre or genre preference in spoiler effects (Johnson & Rosenbaum, 2018; Leavitt & Christenfeld, 2011), they have, so far, not focused on a specific genre. In this project, we specifically look at the effect of spoilers on the enjoyment and suspense of horror films. Horror films are movies wherein the content seems designed to induce a state of fear or terror (Sparks & Sparks, 2000). When studying spoilers, horror is a particularly relevant genre as the enjoyment of horror is linked in part to the uncertainty of not seeing certain violent events coming or not knowing who the next victim will be. Viewers of horror films know to assume that the characters with the most screen time will most likely come to a painful ending (Oliver & Sanders, 2004). When beloved characters are in dangerous situations, this excitement ultimately creates positive emotions (Zillmann et al., 1975). The bigger the threat, the bigger the excitement, and the more pleasure and enjoyment the viewers experience when characters survive the threat. This is due to the slow decay of arousal, and the common misattribution of arousal produced during intense negative emotions (e.g., suspense) to subsequent positive outcomes such as a happy ending (excitation transfer theory; Zillmann, 1980). Spoilers are expected to undermine this uncertainty (Johnson & Rosenbaum, 2015). On the other hand, horror has also been associated with gratifications beyond these sorts of scares, such as the thrill of seeing destruction and gore (e.g., King & Hourani, 2007; Tamborini & Stiff, 1987), which may not be impacted by spoilers at all. Linked to this pleasure in destruction, Carroll (1990) argued that horror films are attractive because they offer images and stories that viewers will not see in their daily lives. Horror movies often show behaviors that violate the norms and values of the society, and this violation of norms arouses the curiosity and fascination of the viewers (Carroll, 1990). According to this line of thinking, spoilers may have little impact on this gratification for viewers. The present study thus focuses on horror movies as a particular case study that could help distinguish alternate factors influencing the relationship between spoilers and audience responses. In particular, regarding the horror genre, these factors are suspense, fluency, reactance, the degree of “spoiling,” and individual differences in viewers. Ó 2019 Hogrefe Publishing

Suspense Described as uncertainty about the desired story outcomes for specific characters, suspense is a prominent hedonic response to narratives (Oliver & Bartsch, 2010), especially those that place characters in mortal danger. Narrative uncertainty heightens arousal and pleasure (Knobloch-Westerwick & Keplinger, 2006; Zillmann et al., 1975), although suspense may be felt even if outcomes are well-known (Gerrig, 1989). Previous experimental research has shown that spoilers slightly reduce self-reported suspense (Johnson & Rosenbaum, 2015). Moreover, a loss of suspense is a rationale most commonly invoked by audiences for spoiler avoidance (Perks & McElrath-Hart, 2016). Fluency If spoilers are likely to have a negative effect on enjoyment by harming suspense, any positive effects are likely transmitted through processing fluency, which is the ease, or fluency, with which a narrative is understood. Spoilers have been shown to benefit fluency in both experimental (Leavitt & Christenfeld, 2013) and survey studies (Ellithorpe & Brookes, 2018), but other evidence has been weak or nonsupportive for fluency as a mediator of spoiler effects (Daniel & Katz, 2018; Johnson & Rosenbaum, 2018; Levine et al., 2016). Reactance Spoilers may also elicit reactance, a negative emotional response to the perceived loss of behavioral freedom when others impose on the individual (Reinhart, Marshall, Feeley, & Tutzauer, 2007). Reactance is relevant for spoilers, as they may be perceived to take away the freedom of whether or not a person wants to know what is going to happen in a narrative. To date, one experiment has tested this proposition, and found that reactance, but not enjoyment, was affected by the presence of spoilers (Johnson & Rosenbaum, 2018, Study 2). Transportation and Enjoyment Audience responses to narratives can also culminate in immersion into the story world, accompanied by affective and cognitive evaluations of the entertainment experience. Transportation (Green & Brock, 2000) is one of the mostly widely used conceptualizations of immersion. During transportation, the reader or viewer mentally travels into the story world and experiences affective, cognitive, and imagery-laden engagement with characters, location, and plot. Exposure to spoilers may lead to a decrease in transportation, as the reader already knows what is going to happen and is, subsequently, more likely to think about other things while viewing or reading the story. Spoilers research to date has found mixed evidence: One study found that the cognitive dimension of transportation alone may be harmed by Journal of Media Psychology (2020), 32(1), 14–25

spoilers (Johnson & Rosenbaum, 2015), while other experiments showed no effect (Johnson & Rosenbaum, 2018, Study 1; Rosenbaum & Johnson, 2016), and a survey revealed that spoilers consistent with source material (e.g., book vs. film) had positive effects on transportation (Johnson & Rosenbaum, 2018, Study 3). Enjoyment is a positive response toward entertainment, and perhaps the most widely studied outcome of narrative consumption (Vorderer, Klimmt, & Ritterfeld, 2004). A variety of measures of enjoyment have been used in spoiler research, from single-item (Leavitt & Christenfeld, 2011; Levine et al., 2016) to multi-item (Ellithorpe & Brookes, 2018; Yan & Tsang, 2016) and multidimensional inventories (Johnson & Rosenbaum, 2015, 2018). Enjoyment is shown to follow from greater suspense (Zillmann et al., 1975) as well as processing fluency (Reber, Schwarz, & Winkielman, 2004), although evidence that these variables mediate effects of spoilers on enjoyment has produced mixed support, with some results showing mediation (Ellithorpe & Brookes, 2018; Johnson & Rosenbaum, 2018, Study 3; Leavitt & Christenfeld, 2013) and others showing none (Johnson & Rosenbaum, 2018, Study 1; Levine et al., 2016).

Minor and Major Spoilers With previous studies finding a mix of positive, null, and negative effects on enjoyment and other outcomes, one explanation could be that spoilers vary in terms of what and how much they reveal. Some spoilers work to enhance processing fluency, while others undermine suspense or generate reactance. Yan and Tsang (2016) compared spoilers for narrative outcomes (conclusions) with spoilers for narrative processes (plot points in the middle of the narrative). Participants overestimated the harm to enjoyment from outcome spoilers but underestimated the harm from process spoilers. While both types of spoilers could reduce uncertainty, audiences respond differently to spoilers in different locations in the narrative. Levine et al. (2016) illustrated that a blunt delivery of spoilers was more likely to harm enjoyment: In other words, spoilers that are perceived to give away more of the plot could be more detrimental. Moreover, Johnson and Rosenbaum (2018, Study 3) found that television spoilers that were consistent with the source material (i.e., a book series) enhanced fluency, whereas spoilers that were inconsistent with the books harmed enjoyment. Similarly, we expect that more modest, incremental spoilers help readers and viewers follow the plot, whereas revealing major plot developments or other deviations from audience expectations are expected to disorient the audience and harm processing fluency. Within the horror context, we propose that minor spoilers are those that disclose suspenseful scenes and scares,

Journal of Media Psychology (2020), 32(1), 14–25

B. K. Johnson et al., Minor and Major Horror Spoilers

whereas major spoilers divulge key plot twists, conclusions, and revelations. Revealing outcomes via a major spoiler is unlikely to benefit fluency as it may still be unclear or disorienting as to how the narrative arrives at this conclusion, whereas minor spoilers about details and the progress of the story could help guide the viewer by placing markers for plot points, allowing them to focus on the story and thus enhancing processing fluency. We thus also anticipate that minor spoilers will enhance transportation by facilitating focus on, and immersion into, the story. However, minor and major spoilers alike may hinder suspense and produce reactance, ultimately harming enjoyment. In addition to comparing the size of the spoiler and significance of the scene that is spoiled, this study also uses a design that accounts for the order of spoilers: where they fall in a typical three-act narrative arc, that is, the beginning, middle, and end (Aristotle, 335 BCE/1926) or the setup, complications, and resolution (Trottier, 2014). Story arcs vary in terms of where major plot points occur, and therefore this study will take this variation into account (Knobloch, Patzig, Mende, & Hastall, 2004; Reagan, Mitchell, Kiley, Danforth, & Dodds, 2016). Accordingly, we divide the stimuli films for this study into three acts: beginning (Act 1), middle (Act 2), and end (Act 3). In the present study design, all Act 1 spoilers are minor, all Act 3 are major, but Act 2 spoilers are either minor or major, reflecting differences in real-world plot structures. Because both types of spoilers will take away the plot-driven suspense and surprise that attracts fans to horror movies, we therefore propose the following: Hypothesis 1 (H1): Both minor and major spoilers will diminish enjoyment. Hypothesis 2 (H2): Reduced suspense will mediate spoiler effects on enjoyment. Hypothesis 3 (H3): Increased reactance will mediate spoiler effects on enjoyment. By contrast, we predict that transportation will be differentially sensitive to the size of the plot point that is spoiled. Minor spoilers allow the viewers to focus on guessing about and anticipating the big twist that still awaits them at the end, and they allow the viewers to focus on other parts of the stories that otherwise could be ignored (cf. Hassoun, 2013), whereas major spoilers could introduce further uncertainty if, for example, the narrative’s path to reaching that reveal is unclear (Johnson & Rosenbaum, 2018). Hypothesis 4 (H4): Minor spoilers will have a positive effect on transportation, and major spoilers will have a negative effect on transportation.

Ó 2019 Hogrefe Publishing

B. K. Johnson et al., Minor and Major Horror Spoilers

Hypothesis 5 (H5): Fluency will mediate the effects of minor and major spoilers on transportation.

Individual Differences Personality traits such as NFC and NFA can have an effect on whether narrative spoilers enhance or diminish preferences and enjoyment (Rosenbaum & Johnson, 2016). Additionally, individual preferences for a format or genre may also moderate the effects of spoilers (Rosenbaum & Johnson, 2016). These findings have not yet been applied to spoilers and the horror genre, although the traits are highly relevant to this genre. Need for Cognition People who relish cognitive activities that take considerable effort are said to be high on need for cognition (Cacioppo & Petty, 1982). These individuals prefer to think deeply about narratives, want to understand a story, and enjoy guessing plot outcomes (Knobloch-Westerwick & Keplinger, 2008). Initial research findings showed that individuals low on NFC, that is, who do not care for abstract thinking or deep thought, were found to have a selective preference for (Rosenbaum & Johnson, 2016) and enjoyment of (Levine et al., 2016) literary spoilers. Follow-up studies by Johnson and Rosenbaum (2018), however, found this interaction did not translate to television/film spoilers. NFC could be argued to be quite relevant for the horror context, as individuals high on this trait may be interested in anticipating plot points and thinking about the twists and turns of the story, while individuals low on NFC could benefit from spoilers as they ensure less cognitive effort is needed to follow the story. Therefore, we predict that: Hypothesis 6 (H6): Spoilers will enhance enjoyment for low-NFC individuals. Need for Affect Need for affect refers to the desire or motivation to seek out and experience emotions (Maio & Esses, 2001). Individuals high on NFA tend to pursue strong emotions; individuals low in NFA would rather avoid strong emotions. Rosenbaum and Johnson (2016) found that the enjoyment of spoiled stories was influenced by NFA, with high-NFA individuals enjoying unspoiled short stories more. Presumably, unspoiled stories allow the reader to experience the arousal that uncertain stories evoke more intensely. Additional research revealed mixed evidence for the role of NFA in spoilers and enjoyment (Johnson & Rosenbaum, 2018), where high-NFA individuals expected less enjoyment and had lower intentions to view a film that was explicitly spoiled in advance. Higher NFA was also associated with Ó 2019 Hogrefe Publishing

more processing fluency in response to exposure to bookconsistent Game of Thrones spoilers, and – surprisingly – more enjoyment when it came to how spoiled viewers perceived themselves to be. However, spoiler effects were not moderated by NFA in a separate study that presented movie and television clips (Johnson & Rosenbaum, 2018). More specific to the horror genre, findings show higher levels of NFA are associated with positive appraisals of negative emotions when viewing horror films (Bartsch, Appel, & Storch, 2010). These so-called meta-emotions (reflective feelings about having experienced emotion) show that enjoyment can be produced by having experienced intense, negative, or ambivalent feelings during exposure. This helps explain why fear, disgust, and suspense can translate into positive experiences for viewers of horror films (Bartsch et al., 2010). These emotions are more valuable for high-NFA individuals; as such, they should value the shock and fear that comes from more unexpected frights (i.e., unspoiled horror). Hypothesis 7 (H7): Spoilers will harm enjoyment for high-NFA individuals. Fandom While research on the impact of spoilers on stories from favored genres has produced contradictory results (Johnson & Rosenbaum, 2018; Leavitt & Christenfeld, 2013; Rosenbaum & Johnson, 2016), when it comes to horror movies, we propose that spoilers interact negatively with fandom. As the genre is seen as attractive by many because of the uncertainty about upcoming scares and plot developments, fans of the horror genre may not enjoy spoiled films (Carroll, 1990). By contrast, individuals who do not like the horror genre may enjoy more certainty in general or find spoilers helpful for managing their expectations and emotions when watching a suspenseful or scary film. Hypothesis 8 (H8): Spoilers will harm enjoyment for individuals who are fans of horror films.

Method An online experiment was conducted with a convenience sample of participants recruited from social networks and online forums for film and horror film fans. A total of 321 individuals completed the questionnaire, but seven participants were screened out for spending less than 10 min on the study, resulting in a final sample of N = 314. The sample comprised 52.5% women, with a mean age of 28.51 years (SD = 9.91). Participants were randomly assigned to one Journal of Media Psychology (2020), 32(1), 14–25

of four horror films (Saw, The Descent, You’re Next, and Insidious). They were presented with the synopsis for the respective film taken from https://www.imdb.com, and then viewed three scenes from the film, each preceded by a short introduction that provided basic context as well as the experimental manipulations. Each film featured scenes from the beginning of the film that set up a problem (Act 1), from the middle of the film that introduce a twist or revelation that raised the stakes (Act 2), and from the climax of the film that provided a resolution to the major action (Act 3). Responses to the film were measured with short items after each scene as well as longer inventories after the final scene.

Design and Manipulation Films were selected based on a pretest (see https://osf.io/ xf96k/) that confirmed that short texts, written for three different scenes from each film, were successfully perceived as spoilers or nonspoilers. The pretest also identified Saw and The Descent as having minor–minor–major arcs, and You’re Next and Insidious as minor–major–major arcs. Minor scenes typically consisted of frights or small twists, while major scenes portrayed consequential twists, reveals, or confrontations. For example, the plot point for the second scene from The Descent was perceived as a minor plot point. The scene shows an incident that, while shocking, does not necessarily explain or connect to the larger questions of where the monstrous threat comes from or how it will be escaped. The unspoiled introduction for this scene read: “After one of the friends, Juno, gets attacked by a creature, she fights back and kills it. She then turns around, believing she hears yet another creature approaching her.” The spoiled introduction read: “After one of the friends, Juno, gets attacked by a creature, she fights back and kills it. She then turns around, accidently stabbing her friend Beth in the neck, believing she was a creature.” By contrast, the second scene for You’re Next was perceived as major because it provides a critical revelation about the events taking place. The unspoiled text read: “After Erin is attacked, she hides in the bathroom, hearing a family member talk to the assassins who came to kill them all off. Erin then quickly begins to learn why the family was targeted.” The spoiled version read: “After Erin is attacked, she hides in the bathroom, hearing one of the sons, Felix, and his girlfriend, Journal of Media Psychology (2020), 32(1), 14–25

B. K. Johnson et al., Minor and Major Horror Spoilers

Zee, talking with the hired assassins, revealing that they hired them to kill the family to get the inheritance.” Each scene was extracted as a video clip, ranging in length from 122 to 226 s, M = 173.67, SD = 32.20. Spoilers were assigned independently for every scene, so that a participant was randomly assigned to a spoiled or unspoiled text for each individual scene. This resulted in a 2 [first act (un) spoiled] 2 [second act (un)spoiled] 2 (second act is minor/major) 2 [third act (un)spoiled] between-subjects design.

Measures Single-Item Perceptions Immediately after each scene, participants responded to six short questions to capture spontaneous responses to (un) spoiled scenes. From 1 = strongly disagree to 7 = strongly agree, participants indicated if: “I liked the film scene,” M = 4.25, SE = 0.09; “The scene was suspenseful,” M = 4.58, SE = 0.08; “It was fun to watch,” M = 3.96, SE = 0.09; “I was scared,” M = 2.99, SE = 0.09; “The film scene was shocking,” M = 3.83, SE = 0.08; and “I felt transported into the film,” M = 3.84, SE = 0.09. Suspense A 6-item scale (Krakowiak & Oliver, 2012), for example, “I felt excited while watching the story,” 1 = strongly disagree to 7 = strongly agree, was reliable, α = .863, M = 4.44, SD = 1.36. Processing Fluency A 4-item scale (Johnson & Rosenbaum, 2018), for example, “This narrative was easy to follow,” 1 = strongly disagree to 7 = strongly agree, was reliable, α = .859, M = 5.25, SD = 1.12. Reactance A 4-item anger arousal subscale measured reactance (Dillard & Shen, 2005), for example, “I felt irritated after being exposed to details about the film,” 1 = strongly disagree to 7 = strongly agree, α = .945, M = 2.75, SD = 1.45. Transportation A 12-item measure (Green & Brock, 2000), for example, “I had a vivid mental image of the characters,” 1 = strongly disagree to 7 = strongly agree, was reliable, α = .797, M = 3.42, SD = 0.95. Enjoyment The 3-item fun scale (Oliver & Bartsch, 2010), for example, “It was fun for me to watch this film,” 1 = strongly disagree to 7 = strongly agree, was reliable, α = .959, M = 4.10, SD = 1.83. Ó 2019 Hogrefe Publishing

B. K. Johnson et al., Minor and Major Horror Spoilers

Table 1. Zero-order correlations among key study variables Variable

1. Suspense 2. Processing fluency

.171**

3. Reactance

.103+

.217***

4. Transportation

.724***

.219***

.057

5. Enjoyment

.628***

.367***

.166**

6. NFC

.007

.217***

.105+

.054

.048

7. NFA

.060

.018

.054

.040

.004

.228***

8. Horror fandom

.117*

.207***

.075

.180**

.339***

.113*

.586***

.017

Note. N = 314. NFC = need for cognition. NFA = need for affect. p < .10. *p < .05. **p < .01. ***p < .001.

Need for Cognition A 17-item scale (Cacioppo, Petty, & Kao, 1984), for example, “I would prefer complex to simple problems,” 1 = strongly disagree to 7 = strongly agree, was reliable, α = .845, M = 4.73, SD = 0.79.

Horror Fandom Participants indicated three favorite film genres from a list of 22 (from https://www.imdb.com). In total, 89 (28.34% of the sample) individuals listed horror as a favorite. Pointbiserial correlations with horror fandom, and other raw correlations among key study variables, appear in Table 1.

previously seen the film. Narratives consumed in the past can still be spoiled during a second viewing (cf. Carroll, 1996); indeed, previous experimental and survey studies show little difference in (the lack of) spoiler effects between familiar and unfamiliar viewers (Johnson & Rosenbaum, 2018). Next, respondents were assigned to a film and asked to turn on their audio. The respondent randomly saw either a spoiled or unspoiled introductory text for each of the three film scenes they viewed. Between each scene, they were asked to complete single-item questions evaluating enjoyment, suspense, and transportation for each individual scene. After all three scenes from the film had been viewed, respondents completed questions about their responses to the film, individual differences, and demographics. Finally, respondents were debriefed on the study’s purpose (spoilers’ effects) and were asked not to disclose the purpose to others. The procedure is depicted in Figure 1.

Procedure

Data Analysis

Respondents accessing the study were shown a short introduction and consent form. They were not told the study was about spoilers but were warned it was about scary films. Respondents were cautioned that “some scenes may contain intense violence, blood, or flashing lights” and that they were free to stop participating at any time. The first part of the questionnaire asked about general film viewing habits and whether participants had seen or heard of any of the four films used in the study. The sample was balanced in its familiarity with the titles (63.7% had seen and 94.6% heard of Saw, 47.1% had seen and 78.0% heard of Insidious, 26.8% had seen and 47.1% heard of The Descent, and 15.9% had seen and 36.0% heard of You’re Next). Participants who had seen or heard of movies were not removed from the sample for hypothesis testing. However, removing cases (n = 135) who had previously seen their assigned film did not substantively alter the results reported here. Any potential effects of spoilers were not heightened or uniquely emergent among participants who had not

To examine the main effects of minor and major spoilers (and corresponding narrative arc), 2 2 2 2 between-subjects ANOVAs tested effects on enjoyment and transportation, as well as suspense, processing fluency, and reactance. Full factorial ANOVAs were conducted on the experimental design, to account for their presentation in combination with each other, despite just a single interaction of interest: second act (un)spoiled second minor/major. Alternative ANOVAs were also administered with only main effects (no factorial interactions) in the model, which produced substantively similar results. While the study design makes use of repeated presentation of spoilers and scenes, the independent manipulation of spoilers for each scene facilitated a between-subjects design rather than a within-subjects design. Next, regression analyses were used to test whether suspense, processing fluency, and reactance mediated effects on enjoyment and transportation, and whether NFC, NFA, and horror attitudes moderated effects on enjoyment

Need for Affect A 10-item measure (Appel, Gnambs, & Maio, 2012), for example, “I feel that I need to experience strong emotions regularly,” 1 = strongly disagree to 7 = strongly agree, was reliable, α = .799, M = 4.83, SD = 0.90.

Ó 2019 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1), 14–25

B. K. Johnson et al., Minor and Major Horror Spoilers

qualified by whether it was a minor or major spoiler. Finally, the between-scene measures were tested with a MANOVA for each of the three acts. Despite the use of repeated measures, the randomized manipulation of spoiled/unspoiled for each scene necessitated the use of three MANOVAs (one for each scene) rather than a within-subjects ANOVA. Additional measures of narrative engagement, sensation seeking, attitudes toward spoilers, and frequency of film viewing were collected but were not considered further. The study design, including predictions and power analysis, was pre-registered1 at https://AsPredicted.org/k6dmi.pdf. The questionnaire, data, and syntax are available at https://osf.io/xf96k/.

Results

Figure 1. Study design and flow of procedure.

and transportation. These models were tested using the PROCESS macro (Hayes, 2013), which allows for bootstrapping mediation tests and probing interactions. Mediation models (bootstrapped from 10,000 samples) tested each experimental factor as the independent variable, enjoyment or transportation as the dependent variable, and suspense, processing fluency, and reactance as parallel mediators, while controlling for other experimental factors. Moderation analyses tested each individual difference as moderator while controlling for the other two individual differences. Additionally, for both mediation and moderation, the effect of the spoiler for Scene 2 was also 1

Testing the prediction that minor and major spoilers for a horror film would produce less enjoyment (H1) as well as effects on transportation (H4) and mediator variables, between-subjects ANOVAs indicated a general lack of main effects of spoilers on enjoyment, transportation, suspense, processing fluency, or reactance. There was a statistically significant effect of spoilers in the second scene on fluency, F(1, 298) = 4.19, p = .042, ηp2 = .014. Spoiled second scenes produced less processing fluency, M = 5.14, SD = 1.27, than unspoiled second scenes, M = 5.37, SD = 0.92. This was not moderated by the magnitude of the spoiler, as the interaction between spoiler and narrative arc was not a significant predictor of processing fluency, F(1, 298) = 0.353, p = .55, ηp2 = .001. Likewise, there was no effect of this interaction on enjoyment, transportation, suspense, or reactance. There was a main effect of narrative arc on processing fluency, F(1, 298) = 4.89, p = .028, ηp2 = .016, such that films with major plot points in the second scene produced more fluency, M = 5.40, SD = 1.00, compared with films with minor plot points in the second scene, M = 5.08, SD = 1.23. H1 and H4, for overall effects of spoilers on enjoyment and transportation as dependent variables, were not supported. Cell means for these ANOVAs appear in Table 2. Mediation analyses examined mediating roles for suspense, reactance, and processing fluency (H2, H3, H5), yet found little to no indirect effects of any spoiler manipulations on enjoyment or transportation, considering suspense, processing fluency, and reactance as parallel mediators. Fluency mediated an effect of spoilers in Scene

Deviations from preregistration should be noted. The initial sample size fell short of the preregistered goal, n = 221, owing to time constraints, and therefore an additional 93 cases were collected at reviewer request in early 2018. Repeated-measures ANOVA was not deemed appropriate for the single-item between-scene perceptions, given that the presence of spoilers was manipulated independently for each scene; thus, MANOVAs were used for each act instead.

Journal of Media Psychology (2020), 32(1), 14–25

Ó 2019 Hogrefe Publishing

B. K. Johnson et al., Minor and Major Horror Spoilers

Table 2. Cell means for ANOVAs Stimulus

Cell size N

Enjoyment M (SD)

Transportation M (SD)

Suspense M (SD)

Fluency M (SD)

Reactance M (SD)

Unspoiled

160

4.23 (1.71)

3.48 (0.90)

4.56 (1.29)

5.31 (1.06)

2.74 (1.45)

Spoiled

154

3.96 (1.94)

3.36 (0.99)

4.30 (1.42)

5.18 (1.19)

2.76 (1.45)

Unspoiled

3.97 (1.87)

3.56 (0.96)

4.43 (1.29)

5.25 (1.05)

2.66 (1.30)

Spoiled

4.18 (1.84)

3.49 (1.04)

4.57 (1.40)

4.94 (1.35)

2.85 (1.38)

Unspoiled

4.11 (1.87)

3.39 (0.95)

4.39 (1.43)

5.47 (0.78)

2.95 (1.59)

Spoiled

4.10 (1.77)

3.29 (0.84)

4.36 (1.31)

5.33 (1.18)

2.55 (1.48)

Unspoiled

160

4.14 (1.76)

3.39 (0.89)

4.34 (1.35)

5.29 (1.11)

2.75 (1.49)

Spoiled

154

4.05 (1.90)

3.46 (1.01)

4.54 (1.36)

5.20 (1.14)

2.76 (1.41)

Act 1: Minor

Act 2: Minor

Act 2: Major

Act 3: Major

2 on transportation, b = 0.020, SE = 0.014, 95% CI [ 0.058, 0.0004]. The presence of a major plot point in Scene 2 led to more fluency, which mediated effects on enjoyment, b = 0.130, SE = 0.062, 95% CI [0.024, 0.267], and transportation, b = 0.029, SE = 0.018, 95% CI [0.005, 0.078]. Yet, the presence of a spoiler in Scene 2 did not – as hypothesized – interact with the major versus minor size of the scene to predict mediation via fluency; spoilers and plot size affected fluency independently (and in opposing directions). As such, mediation hypotheses H2, H3, and H5 were not supported. Regression analyses testing the moderating roles of traits (H6, H7, H8) also yielded minimal effects. Minor spoilers were found to have a marginally positive relationship with enjoyment when the NFA was higher (opposite to the directional prediction). A spoiler in Scene 1 yielded a negative effect on enjoyment when NFA was low. The interaction term was not statistically significant, b = 0.40, SE = 0.22, p = .074, ΔR2 = .009. Yet Johnson–Neyman probing revealed that the interaction between NFA and first-act spoilers was characterized by a statistically significant negative effect on enjoyment for those low on NFA (< 4.77, 42.12% of the sample). Similarly, a three-way interaction fell short of significance for NFA, second-act spoilers, and narrative arc, b = –0.86, SE = 0.44, p = .052, ΔR2 = .011. The interaction was characterized by a nonsignificant negative effect for those very low on NFA, as well as a positive effect of spoilers for those high on NFA, which was not significant for those individuals, with +1 SD (> 5.73), but instead only significant for extremely high NFA individuals (> 6.23, 4.18% of the sample), but only when the spoiler was for a minor scene (i.e., for Saw or The Descent). Major spoilers were not moderated by NFA. Given the failure to meet statistical significance, and a pattern of results in the opposite direction, H7 was not supported. No moderating effects were evident for NFC or horror fandom; thus, H6 and H8 were not supported. Ó 2019 Hogrefe Publishing

Additionally, transportation was examined as a preregistered alternative to enjoyment as dependent variable in moderation models. No significant interactions emerged between spoilers and the moderators of NFC or horror fandom. For NFA, the three-way interaction with Scene 2 spoiler and Scene 2 size was significant, b = 0.64, SE = 0.24, p = .007, ΔR2 = .023. Johnson–Neyman probing revealed a negative effect of the spoiler on transportation for low-NFA individuals (< 3.77, 10.29% of the sample), and a positive effect of the spoiler on transportation for high-NFA (> 5.70, 15.11% of the sample) individuals, but only when the scene was minor (Figure 2). This result mirrors the pattern of results for enjoyment of minor spoilers in the second act. Finally, in another preregistered exploration, we examined single-item measures (enjoyment, suspense, fun, scare, shock, and transportation) that were administered after each scene. A MANOVA was constructed for each of the three film scenes. No effects were evident for Scene 1, ΛLH = .012, F(6, 306) = 0.59, p = .73, ηp2 = .012, Scene 2, ΛLH = .010, F(6, 307) = 0.53, p = .79, ηp2 = .010, or Scene 3, ΛLH = .010, F(6, 307) = 0.50, p = .81, ηp2 = .010, nor did any individual univariate tests emerge as significant. Likewise, the items were unaffected by the size of Scene 2’s spoiler (i.e., the interaction between Scene 2 spoiler and narrative arc), ΛLH = .029, F(6, 305) = 1.46, p = .19, ηp2 = .028.

Discussion This study advances the state of research into effects of narrative spoilers on enjoyment by employing spoilers of different magnitudes at different places in the narrative and by testing proposed mechanisms and conditional factors in the highly relevant context of horror films. As with the bulk of the existing research into spoilers, effects on enjoyment are largely absent, and very small and qualified when they do exist. Journal of Media Psychology (2020), 32(1), 14–25

B. K. Johnson et al., Minor and Major Horror Spoilers

Figure 2. Three-way interaction between spoilers, narrative arc, and need for affect (NFA).

None of our preregistered hypotheses were supported. No significant main or indirect effects of spoilers were evident in the present study. Spoilers in the second scene produced more fluency, as did the presence of a major plot point in that scene. However, the interaction of spoiler and narrative arc (such that spoilers were minor) did not impact processing fluency and subsequent transportation. Yet, there was evidence that spoilers might have some influence on enjoyment and transportation. Contrary to expectations, minor spoilers in the first act or second act marginally harmed enjoyment for individuals low on NFA. High-NFA individuals showed positive effects of minor spoilers in the second act (but not the first act). These interactions fell short of statistical significance, yet this same three-way interaction did produce an effect on transportation, such that low-NFA participants (those who do not value emotional experiences) showed negative effects of spoilers for minor plot points in Act 2. We hypothesized (after Rosenbaum & Johnson, 2016), that high-NFA individuals and horror fans would value the uncertainty and arousal provided by the absence of spoilers. Instead, high-NFA individuals reacted slightly positively to spoiled scenes with minor plot points, and lowNFA individuals reacted negatively to these minor spoilers. The results correspond to those in a recent survey that showed NFA moderated a positive effect of perceived spoiler exposure on enjoyment of a fantasy television series (Johnson & Rosenbaum, 2018). Our findings, however, suggest spoilers may play a different role in the horror genre than other types of media. The expectation of a fright or small twist may heighten enjoyment through anticipation of the scare. To that end, we Journal of Media Psychology (2020), 32(1), 14–25

cautiously offer a post hoc interpretation of the results in which minor spoilers in particular can be beneficial as they direct attention to a coming fright, increasing viewer anticipation. This state of anticipation could heighten the viewer’s concern for the well-being of beloved characters in peril (e.g., Tamborini, Stiff, & Heidel, 1990), even if they are more certain of the character’s fate. Anxious waiting could be pleasurable for high-NFA individuals. However, the small effect size cautions against drawing robust claims about this explanation. In contrast, the lack of moderated effects of spoilers for major plot points, and the general lack of main effects of spoilers, adds to growing evidence that spoilers produce small and inconsistent or qualified effects. Even the moderated effects of minor spoilers seen here yielded very small effect sizes, in keeping with spoiler research to date (Johnson & Rosenbaum, 2018; Leavitt & Christenfeld, 2011). Thus, it is not sufficient to point out that spoilers are beneficial for some people and harmful for others (Rosenbaum & Johnson, 2016), without further qualification. As Carroll (1996) observed with regard to re-watching of narratives, the paradox of suspense is that we can feel uncertain even when we know the ending of a storyline, and we can generally re-experience the thrill of a narrative’s twists and turns as if we had come to it with fresh eyes and ears. The present study also failed to replicate processing fluency as a mediator of the impact of spoilers. Existing evidence for this mechanism is mixed (Ellithorpe & Brookes, 2018; Johnson & Rosenbaum, 2018; Leavitt & Christenfeld, 2013; Levine et al., 2016). While fluency produces transportation and enjoyment, the influence of spoilers on fluency seems tenuous, requiring further examination. Ó 2019 Hogrefe Publishing

B. K. Johnson et al., Minor and Major Horror Spoilers

These minimal effects are in stark contrast to common layperson beliefs about the harm of spoilers, and suggest that media consumers, even for horror films, should not overly worry about spoilers in online discussions, promotional materials, or critics’ reviews. Yet why do these beliefs about the harmful effect of spoilers persist, and why did they develop in the first place? One promising suggestion is that spoilers are disliked because they violate the user’s autonomy. A previous experiment illustrated that psychological reactance was activated by spoilers (Johnson & Rosenbaum, 2018). However, the present study did not replicate this effect on reactance. Another promising explanation for the discrepancy between spoiler beliefs and spoiler effects could be affective forecasting (Yan & Tsang, 2016). Readers and viewers may simply fail to account for their ability to find pleasure in narratives despite spoilers or other hindrances. Emerging norms and discourse around narrative entertainment, especially as it takes shape online, may play an important role. Limitations of the present approach should be acknowledged. Experimentation can control for the delivery of spoilers, including their nature and location, but does not perfectly reproduce in vivo reception of spoilers, in which peers, journalists, or online users may quickly, inadvertently reveal spoilers for long-anticipated installments or new releases. Survey research is an important corrective to these limitations (Ellithorpe & Brookes, 2018; Johnson & Rosenbaum, 2018), but it is prone to other biases such as poor recall, natural confounds, and socially desirable reporting. Subsequently, it is possible that the shortcomings of both methods mean that the impact of spoilers is never fully captured in research. A final limitation is the use of scenes rather than entire films; yet using multiple scenes does improve on previous studies that used single scenes, vignettes, or short stories.

Conclusion The present study provides directions for future research. While our study used various nuanced measures, future experimental and survey research could employ measures that identify actual beliefs about what qualifies as (un)certain (cf. Ellithorpe & Brookes, 2018) and continuous responses such as physiological indicators of affect. While the horror context is promising in terms of its relevance for spoilers, other genres and media should be explored as well, such as the implications of spoilers for frightening and violent interactive and immersive media (Lynch & Martins, 2015). The modest effects found in the present study, in which spoilers seemed to enhance the thrill of anticipating scares among individuals who value intense emotions, should also be investigated further. Ó 2019 Hogrefe Publishing

In short, while the present study shows spoilers appear to have much weaker and conditional effects than widely believed, they could have positive potential for focusing some viewers on the impending twists, turns, and scares delivered by highly arousing media. Waiting for a scare that one knows is coming can make for an enjoyable moment.

Research Transparency Statement The authors are willing to share their data, analytics methods, and study materials with other researchers.

References Appel, M., Gnambs, T., & Maio, G. R. (2012). A short measure of the need for affect. Journal of Personality Assessment, 94, 418– 426. https://doi.org/10.1080/00223891.2012.666921 Aristotle. (1926). Poetics (1450b27), [W. H. Fyfe, Trans., Originally published c. 335 BCE]. Retrieved from http://www.perseus. tufts.edu/hopper/text?doc=Aristot.+Poet.+1450b Bartsch, A., Appel, M., & Storch, D. (2010). Predicting emotions and meta-emotions at the motives: The role of the need for affect in audiences’ experiences of horror and drama. Communication Research, 37, 167–190. https://doi.org/10.1177/ 0093650209356441 Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116–131. https://doi.org/10.1037/0022-3514.42.1.116 Cacioppo, J. T., Petty, R. E., & Kao, C. F. (1984). The efficient assessment of need for cognition. Journal of Personality Assessment, 48, 306–307. https://doi.org/10.1207/s15327752jpa4803_ 13 Carroll, N. (1990). The philosophy of horror: Or, paradoxes of the heart. London, UK: Routledge. Carroll, N. (1996). The paradox of suspense. In P. Vorderer, H. J. Wulff, & M. Friedrichsen (Eds.), Suspense: Conceptualization, theoretical analysis, and empirical explorations (pp. 71–90). Mahwah, NJ: Erlbaum. https://doi.org/10.4324% 2F9780203811252 Daniel, T. A., & Katz, J. S. (2018). Spoilers affect the enjoyment of television episodes but not short stories. Psychological Reports. Advance online publication. https://doi.org/10.1177/ 0033294118793971 Dillard, J. P., & Shen, L. (2005). On the nature of reactance and its role in persuasive health communication. Communication Monographs, 72, 144–168. https://doi.org/10.1080/ 03637750500111815 Ellithorpe, M. E., & Brookes, S. E. (2018). I didn’t see that coming: Spoilers, fan theories, and their influence on enjoyment and parasocial breakup distress during a series finale. Psychology of Popular Media Culture, 7, 250–263. https://doi.org/10.1037/ ppm0000134 Gerrig, R. J. (1989). Suspense in the absence of uncertainty. Journal of Memory and Language, 28, 633–648. https://doi.org/ 10.1016/0749-596X(89)90001-6 Green, M. C., & Brock, T. C. (2000). The role of transportation in the persuasiveness of public narratives. Journal of Personality and Social Psychology, 79, 701–721. https://doi.org/10.1037/ 0022-3514.79.5.701

Journal of Media Psychology (2020), 32(1), 14–25

Hassoun, D. (2013). Sequential outliers: The role of spoilers in comic book reading. Journal of Graphic Novels and Comics, 4, 346–358. https://doi.org/10.1080/21504857.2013.784202 Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: The Guilford Press. Johnson, B. K., & Rosenbaum, J. E. (2015). Spoiler alert: Consequences of narrative spoilers for dimensions of enjoyment, appreciation, and transportation. Communication Research, 42, 1068–1088. https://doi.org/10.1177/0093650214564051 Johnson, B. K., & Rosenbaum, J. E. (2018). (Don’t) tell me how it ends: Spoilers, enjoyment, and involvement in television and film. Media Psychology, 21, 582–612. https://doi.org/10.1080/ 15213269.2017.1338964 King, C. M., & Hourani, N. (2007). Don’t tease me: Effects of ending type on horror film enjoyment. Media Psychology, 9, 473–492. https://doi.org/10.1080/15213260701282915 Knobloch, S., Patzig, G., Mende, A. M., & Hastall, M. (2004). Affective news: Effects of discourse structure in narratives on suspense, curiosity, and enjoyment while reading news and novels. Communication Research, 31, 259–287. https://doi.org/ 10.1177/0093650203261517 Knobloch-Westerwick, S., & Keplinger, C. (2006). Mystery appeal: Effects of uncertainty and resolution on the enjoyment of mystery. Media Psychology, 8, 193–212. https://doi.org/ 10.1207/s1532785xmep0803_1 Knobloch-Westerwick, S., & Keplinger, C. (2008). Murder for pleasure: Impacts of plot complexity and need for cognition on mystery enjoyment. Journal of Media Psychology, 20, 117– 128. https://doi.org/10.1027/1864-1105.20.3.117 Krakowiak, K. M., & Oliver, M. B. (2012). When good characters do bad things: Examining the effect of moral ambiguity on enjoyment. Journal of Communication, 62, 117–135. https:// doi.org/10.1111/j.1460-2466.2011.01618.x Leavitt, J. D., & Christenfeld, N. J. S. (2011). Story spoilers don’t spoil stories. Psychological Science, 22, 1152–1154. https://doi. org/10.1177/0956797611417007 Leavitt, J. D., & Christenfeld, N. J. S. (2013). The fluency of spoilers: Why giving away endings improves stories. Scientific Study of Literature, 3, 93–104. https://doi.org/10.1075/ ssol.3.1.09lea Levine, W. H., Betzner, M., & Autry, K. S. (2016). The effect of spoilers on the enjoyment of short stories. Discourse Processes, 53, 513–531. https://doi.org/10.1080/0163853X.2016.1141350 Lynch, T., & Martins, N. (2015). Nothing to fear? An analysis of college students’ fear experiences with video games. Journal of Broadcasting & Electronic Media, 59, 298–317. https://doi.org/ 10.1080/08838151.2015.1029128 Maio, G. R., & Esses, V. M. (2001). The need for affect: Individual differences in the motivation to approach or avoid emotions. Journal of Personality, 69, 583–614. https://doi.org/10.1111/ 1467-6494.694156 Oliver, M. B., & Bartsch, A. (2010). Appreciation as audience response: Exploring entertainment gratifications beyond hedonism. Human Communication Research, 36, 53–81. https://doi. org/10.1111/j.1468-2958.2009.01368.x Oliver, M. B., & Sanders, M. (2004). The appeal of horror and suspense. In S. Prince (Ed.), The horror film (pp. 242–260). New Brunswick, NJ: Rutgers University Press. Perks, L. G., & McElrath-Hart, N. (2016). The television spoiler nuisance rationale. International Journal of Communication, 10, 5580–5597. Reagan, A. J., Mitchell, L., Kiley, D., Danforth, C. M., & Dodds, P. S. (2016). The emotional arcs of stories are dominated by six basic shapes. EPJ Data Science, 5, 31. https://doi.org/10.1140/ epjds/s13688-016-0093-1

Journal of Media Psychology (2020), 32(1), 14–25

B. K. Johnson et al., Minor and Major Horror Spoilers

Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is beauty in the perceiver’s processing experience? Personality and Social Psychology Review, 8, 364–382. https://doi.org/10.1207/s15327957pspr0804_3 Reinhart, A. M., Marshall, H. M., Feeley, T. H., & Tutzauer, F. (2007). Persuasive effects of message framing in organ donation: The mediating role of psychological reactance. Communication Monographs, 74, 229–255. https://doi.org/10.1080/ 03637750701397098 Rosenbaum, J. E., & Johnson, B. K. (2016). Who’s afraid of spoilers? Need for cognition, need for affect, and narrative selection and enjoyment. Psychology of Popular Media Culture, 5, 273–289. https://doi.org/10.1037/ppm0000076 Sparks, G. G., & Sparks, C. W. (2000). Violence, mayhem, and horror. In D. Zillmann & P. Vorderer (Eds.), Media entertainment: The psychology of its appeal (pp. 73–91). Mahwah, NJ: Erlbaum. Tamborini, R., & Stiff, J. (1987). Predictors of horror film attendance and appeal an analysis of the audience for frightening films. Communication Research, 14, 415–436. https://doi.org/ 10.1177/009365087014004003 Tamborini, R., Stiff, J., & Heidel, C. (1990). Reacting to graphic horror: A model of empathy and emotional behavior. Communication Research, 17, 616–640. https://doi.org/10.1177/ 009365090017005003 Topolinksi, S. (2014). A processing fluency-account of funniness: Running gags and spoiling punchlines. Cognition and Emotion, 28, 811–820. https://doi.org/10.1080/02699931.2013.863180 Trottier, D. (2014). The screenwriter’s bible: A complete guide to writing, formatting, and selling your script (6th ed.). Hollywood, CA: Silman-James Press. Vorderer, P., Klimmt, C., & Ritterfeld, U. (2004). Enjoyment: At the heart of media entertainment. Communication Theory, 14, 388–408. https://doi.org/10.1111/j.1468-2885.2004.tb00321.x Yan, D., & Tsang, A. S. L. (2016). The misforecasted spoiler effect: Underlying mechanism and boundary conditions. Journal of Consumer Psychology, 26, 81–90. https://doi.org/10.1016/j. jcps.2015.05.003 Zillmann, D. (1980). Anatomy of suspense. In P. H. Tannenbaum (Ed.), The entertainment functions of television (pp. 133–163). Hillsdale, NJ: Erlbaum. Zillmann, D., Hay, T. A., & Bryant, J. (1975). The effect of suspense and its resolution on the appreciation of dramatic presentations. Journal of Research in Personality, 9, 307–323. https:// doi.org/10.1016/0092-6566(75)90005-7 History Received November 9, 2017 Revision received June 16, 2018 Accepted September 9, 2018 Published online March 8, 2019 Acknowledgments This manuscript is partially based on Angel Udvardi’s master thesis, which was written under the supervision of Benjamin K. Johnson. Open Data All pretest data, the questionnaire, data, and syntax are available at https://osf.io/xf96k/. The study design, including predictions and power analysis, was preregistered at https://AsPredicted.org/ k6dmi.pdf. ORCID Benjamin K. Johnson https://orcid.org/0000-0003-0588-221X

Ó 2019 Hogrefe Publishing

B. K. Johnson et al., Minor and Major Horror Spoilers

Benjamin K. Johnson Department of Advertising University of Florida 1885 Stadium Road Gainesville, FL 32611 USA benjaminkjohnson@ufl.edu

Allison Eden (PhD, 2011) is Assistant Professor of Communication at Michigan State University, USA, specializing in media entertainment research. Her work is interdisciplinary, drawing from communication and media psychology, social psychology, and neuroscience in the areas of media psychology, media entertainment, and media processing. Benjamin K. Johnson (PhD, 2014) is Assistant Professor of Advertising at the University of Florida, previously of Communication Science at Vrije Universiteit Amsterdam, The Netherlands. His research is focused on selective exposure in new media settings, especially as it relates to impression management, social comparison, and selfregulation processes.

Judith E. Rosenbaum (PhD, 2007) is Assistant Professor of Communication and Journalism at the University of Maine, USA. Her research examines how people interact with the media and what factors drive this interaction. Moreover, she is interested in how people’s media use can become a tool to engender individual and social change.

Angel Udvardi (MSc, 2016) is Digital Advertising Consultant at Traffic Builders, where she is responsible for the SEA and social campaigns of various clients. She is interested in all developments of the media landscape and its effect on the consumer.

Ó 2019 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1), 14–25

Erratum Correction to Johnson, Udvardi, Eden, & Rosenbaum, 2020 The article entitled “Spoilers go bump in the night: Impacts of minor and major reveals on horror film enjoyment” by Johnson, B. K., Udvardi, A., Eden, A., & Rosenbaum, J. E. (2020), Journal of Media Psychology, 32, 14–25. https:// doi.org/10.1027/1864-1105/a000252. Advance online publication March 8, 2019 contained an error on p. 14. The authors’ affiliation should read correctly: Benjamin K. Johnson1, Angel Udvardi2, Allison Eden3, and Judith E. Rosenbaum4 1

Department of Advertising, University of Florida, Gainesville, FL, USA 2 Department of Communication Science, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands 3 Department of Communication, Michigan State University, East Lansing, MI, USA 4 Department of Communication and Journalism, University of Maine, Orono, ME, USA

Journal of Media Psychology (2020), 32(1), 26 https://doi.org/10.1027/1864-1105/a000276

We regret any inconvenience or confusion this error may have caused.

Reference Johnson, B. K., Udvardi, A., Eden, A., & Rosenbaum, J. E. (2020). Spoilers go bump in the night: Impacts of minor and major reveals on horror film enjoyment. Journal of Media Psychology, 32, 14–25. https://doi.org/10.1027/1864-1105/a000252

Published online February 25, 2020

Ó 2020 Hogrefe Publishing

Assessment and treatment of Internet addiction

“This excellent book is a pleasure to read. At a time when clinicians are scrambling to learn what they can about the rapidly developing problem of Internet addiction, this book offers them an excellent place to start.” Hilarie Cash, PhD, Chief Clinical Officer and Co-Founder of reSTART Life, PLLC, Fall City, WA – the first residential treatment program for Internet addiction in the US

Daria J. Kuss / Halley M. Pontes

Internet Addiction (Series: Advances in Psychotherapy – Evidence-Based Practice – Volume 41) 2019, iv + 86 pp. US $29.80 / € 24.95 ISBN 978-0-88937-501-7 Also available as eBook This book examines how you can identify, assess, and treat Internet addiction in the most effective manner. Internet use has become an integral part of our daily lives, but at what point does it become problematic? What are the different kinds of Internet addiction? And how can professionals best help clients? This compact, evidencebased guide written by leading experts from the field helps disentangle the debates and controversies around Internet addiction, including social media addiction

www.hogrefe.com

and Internet gaming disorder, and outlines the current assessment and treatment methods. The book presents a 12–15 session treatment plan for Internet and gaming addiction using the method and setting with the best evidence: group CBT. Printable tools in the appendix help clinicians implement therapy. This accessible book is essential reading for clinical psychologists, psychiatrists, psychotherapists, counselors, social workers, teachers, researchers, as well as students and parents.

Start using strengths today! “The GO-TO book for building character.” Martin E. P. Seligman, the founder of positive psychology

Ryan M. Niemiec

Character Strengths Interventions A Field Guide for Practitioners 2018, xx + 300 pp. US $59.00 / € 46.95 ISBN 978-0-88937-492-8 Also available as eBook This book is the epitome of positive psychology: it takes the “backbone” of positive psychology – character strengths – and builds a substantive bridge between the science and practice. Working with clients’ (and our own) character strengths boosts well-being, fosters resilience, improves relationships, and creates strong, supportive cultures in our practices, classrooms, and organizations. This unique guide brings together the vast experience of the author with the science and the practice of positive psychology in such a way that both new and experienced practitioners will benefit. New practitioners will learn about the core concepts of character and signature strengths and how to fine-tune their approach and troubleshoot. Experienced practitioners will deepen their knowledge about advanced topics such as strengths overuse and collisions, hot button issues, morality, and integrating strengths with savor-

www.hogrefe.com

ing, flow, and mindfulness. Hands-on practitioner tips throughout the book provide valuable hints on how to take a truly strengths-based approach. The 24 summary sheets spotlighting each of the universal character strengths are an indispensable resource for client sessions, succinctly summarizing the core features of and research on each strength. 70 evidence-based step-by-step activity handouts can be given to clients to help them develop character strengths awareness and use, increase resilience, set and meet goals, develop positive relationships, and find meaning and engagement in their daily lives. No matter what kind of practitioner you are, this one-of-a-kind field guide is a goldmine in science-based applications. You’ll be able to immediately bring the science of well-being into action!

Original Article

Associations Between Different Aspects of Video Game Play Behavior and Adolescent Adjustment Geert P. Verheijen , William J. Burk, Sabine E. M. J. Stoltz, Yvonne H. M. van den Berg, and Antonius H. N. Cillessen Behavioural Science Institute, Radboud University, Nijmegen, The Netherlands Abstract: Playing video games during adolescence has been linked to both positive and negative outcomes, but it remains unclear which specific aspects of gaming behavior are associated with adjustment. The current study examines the association between distinct aspects of video game play behavior and adolescent adjustment. We focused on five aspects, namely, gaming frequency, the social context, gamer identification, type of game, and motivation for playing. Participants included 705 Dutch adolescents (33.5% female, Mage = 14.07), who completed a survey including self-report and sociometric measures. Results indicated that the frequency of playing games was not associated with adolescent adjustment. However, significant associations did appear for social context, genre of games, and motivation for playing. In addition, the different aspects of gaming showed both beneficial and deleterious relations with adjustment. This research indicates the importance of disaggregating gaming behavior and the necessity to look beyond frequency of play in order to fully understand the impact of gaming on youth. Keywords: adolescence, digital games, video game play behavior, adolescent adjustment

Video games are a major leisure activity for today’s youth. In total, 72% of teens play video games (Lenhart, Smith, Anderson, Duggan, & Perrin, 2015; van Dorsselaer et al., 2016) and this high prevalence has left parents, teachers, and policy makers alike worried about the impact of this pastime on youth. Researchers followed suit by initially focusing on examining the possible negative consequences of playing games. More recently, however, the field is shifting to also consider the possible benefits of video games (Granic, Lobel, & Engels, 2014). Indeed, an increasing number of studies report that playing video games is associated with both positive and negative adjustment for adolescents (Allahverdipour, Bazargan, Farhadinasab, & Moeini, 2010; Durkin & Barber, 2002; Przybylski, 2014). This more nuanced approach to gaming research is an important next step for the field. However, we argue that understanding of gaming effects remains hindered by a predominant focus on the amount of time spent on video games as a main predictor. The aim of the current study was to build upon previous work that emphasizes the multidimensionality of gaming and to disaggregate video game play behavior into distinct aspects. For this, we examined how five aspects of gaming behavior are uniquely associated with separate indicators of adolescent Ó 2019 Hogrefe Publishing

adjustment, such as loneliness, self-esteem, need to belong, fear of negative evaluation, and aggression.

Looking Beyond Frequency of Play Traditionally, gaming research has focused on frequency of play (e.g., screen time) for two main reasons. The first comes from displacement theory, which posits that the time spent playing video games displaces activities that are considered more valuable for youth, such as studying and social interactions. In this way, the number of hours spent playing video games is assumed to have a negative impact on adjustment (Hofferth, 2010). The second reason for studying frequency is based on the idea that the amount of time spent playing video games provides a suitable proxy that represents all aspects of gaming behavior. Measuring the amount of time spent gaming includes other underlying aspects inherent to gaming behavior, which might be more directly related to adjustment, such as satisfaction of basic psychological needs (Przybylski, Rigby, & Ryan, 2010). While these are both valid approaches, we emphasize that gaming is a complex behavior and research should consider gaming aspects besides frequency. Only considering the amount of time spent playing games runs the risk of Journal of Media Psychology (2020), 32(1), 27–39 https://doi.org/10.1027/1864-1105/a000253

missing or misinterpreting findings. This is illustrated by the fact that associations between time spent playing video games and indicators of adolescent adjustment are often inconsistent in the literature. For instance, several studies found that gaming is associated with increased depressive symptoms (Casiano, Kinley, Katz, Chartier, & Sareen, 2012; Maras et al., 2015; Romer, Bagdasarov, & More, 2013), while other studies were unable to confirm this association (Ferguson, 2015; Merritt, LaQuea, Cromwell, & Ferguson, 2016; Primack, Swanier, Georgiopoulos, Land, & Fine, 2009). Frequency of gaming has been linked to lower sociability in some studies (Griffiths, 2010; Heim, Brandtzæg, Kaare, Endestad, & Torgersen, 2007; Padilla-Walker, Nelson, Carroll, & Jensen, 2010), but not in others (Brooks, Chester, Smeeton, & Spencer, 2016; Wack & TantleffDunn, 2009; Willoughby, 2008). Lower self-esteem is sometimes found to be related to higher video game use (Colwell & Payne, 2000; Witt, Massman, & Jackson, 2011), but not in all studies (Colwell & Kato, 2003; Kowert, Vogelgesang, Festl, & Quandt, 2015). Finally, measures of general well-being have been associated negatively with high frequency of gaming (Brooks et al., 2016; Przybylski, 2014; Yang, Helgason, Sigfusdottir, & Kristjansson, 2013), but not always (Willoughby, 2008). There are several possible explanations for these inconsistent results. First, most studies of gaming frequency on well-being find only small effects sizes (Ferguson, 2015), and the field may suffer from power issues. Alternatively, some researchers have argued for a curvilinear relationship between frequency and well-being, where moderate gamers show better adjustment than both non-gamers and highfrequency gamers (Allahverdipour et al., 2010; Durkin & Barber, 2002; Przybylski, 2014). Finally, it is possible that separate positive and negative underlying aspects of gaming cancel each other out when only the general proxy of frequency is measured. To counteract these issues, research should look beyond frequency and focus on the underlying aspects of video game play behavior in adolescent adjustment. The importance of looking beyond frequency of play has been argued before (Gentile, 2011; Johnson & Sweetser, 2013) and is illustrated in existing research. For instance, Shen and Williams (2011) found that time spent playing massive multiplayer online (MMO) games had a small negative impact on psychosocial functioning, but other aspects, such as the motivation for playing and with whom the game was played, had larger associations. Similarly, Vella, Johnson, and Hides (2013) found no impact of frequency of play, but found that with whom games were being played and experiences during play were positively associated with a measure of emotional, social, and psychological well-being. The current study expands on these findings by examining the relation between adolescent adjustment and five Journal of Media Psychology (2020), 32(1), 27–39

G. P. Verheijen et al., Video Games and Adolescent Adjustment

distinct aspects of video gaming behavior: frequency of play, social context, gamer identification, type of game, and motivation for playing. These five aspects where chosen to represent a more complete range of different behaviors associated with video game play and may be distinctly related to adolescent adjustment.

Five Aspects of Video Game Play Behavior We considered the frequency of playing video games first. The number of hours someone spends playing video games is important, as there can be no impact of gaming if no time is spent playing games. Although the relation between frequency and adolescent adjustment has been inconsistent, we hypothesized that frequency shows general negative associations with adjustment owing to displacement effects (Hofferth, 2010). Second, we looked at the social context in which video games are played. Games allow players to engage socially with others in a myriad ways, resulting in varying associations with adjustment. For example, compared with offline play, online play has been shown to bolster life satisfaction among adolescents (Kowert et al., 2015), but also predicts a smaller offline social circle (Kowert, Domahidi, Festl, & Quandt, 2014). Furthermore, the nature of an opponent (i.e., computer, friend, or stranger) has been found to affect physiological arousal and engagement during play (Ravaja et al., 2006) as well as general well-being (Vella et al., 2013). We hypothesized that, compared with playing online with strangers or alone, playing together with friends would be associated with lower loneliness, need to belong, and fear of negative evaluation. Third, we took the gamer identification of the adolescent into account. Only a portion of people who play games actually consider themselves a gamer (De Grove, Courtois, & van Looy, 2015). The social category of gamers is often stereotyped as being unpopular, unattractive, and socially incompetent (Kowert, Griffiths, & Oldmeadow, 2012) and adolescents who strongly identify as gamers might feel stigmatized by these negative connotations (Aarsand, 2012). We therefore explored whether there were any associations between gamer identification and adjustment. Fourth, we looked at the type of games that adolescents play. Games are often classified in genres, such as racing games or first-person shooters. No official classification exists and games are often a mix of different genres, which means mapping specific games onto genres is difficult. Yet some genre-specific associations have been shown. On the basis of earlier research, we hypothesized that genres encouraging social interaction (e.g., social network games) would be associated with lower loneliness, need to belong, and fear of negative evaluation (Wang, 2014). Furthermore, we hypothesized that aggression would be Ó 2019 Hogrefe Publishing

G. P. Verheijen et al., Video Games and Adolescent Adjustment

associated with genres that are either competitive in nature (Adachi & Willoughby, 2016) or typically violent in content (Anderson et al., 2010). While the influence of violent video games on aggression has been brought into question (Ferguson, 2015; Kühn et al., 2018), nearly every metaanalysis has produced small positive effects (Markey, 2015), which is what we expected to replicate. Fifth, we considered a person’s motivation for playing video games. People play games for different reasons, and even the same video game may have a different impact on player well-being based on the motivation for engaging in play. A person who picks up a game with the intention of connecting with other players might have a different experience than someone who uses the same game for relieving stress. The self-determination theory posits that people play games to satisfy their needs for autonomy, competence, and relatedness (Przybylski et al., 2010). Need fulfillment through gaming has been linked to enjoyment, frequency of play, and well-being (Przybylski, Weinstein, Ryan, & Rigby, 2009). On the basis of associations between catharsis and elevated depressive symptoms (Ferguson & Olson, 2013), we hypothesized an overall negative association between catharsis and internalizing behavior. Furthermore, we expected that social motivations for playing games is associated with loneliness, need to belong, and fear of negative evaluation, but did not specify a direction. Social motivations might illustrate a person’s opportunities to play with friends, but also indicate an unmet need for sociability (Kowert et al., 2015).

Current Research Goals The aim of the current study was to illustrate how each of these five aspects may uniquely impact adolescent adjustment. The first goal was to examine the unique contribution of aspects of gaming behavior, beyond frequency of play. We expected an initial negative relation between frequency of play and adolescent adjustment. We further hypothesized that this association would be reduced when other aspects of gaming behavior were included in the analyses, illustrating that the impact of video games on youth will be clarified when frequency is complemented with other aspects of gaming. The second goal was to examine the uniqueness of different aspects of video gaming by investigating their distinct associations with adolescent adjustment. While there were some expectations based on literature, our approach was exploratory. In general, we expected that separate aspects of gaming behavior would 1

uniquely predict different domains of adjustment, and that both positive and negative associations within the same domain of adolescent adjustment would appear. This would illustrate that video games are not a uniform predictor, but that different aspects of gaming have distinct, unique associations with adolescent adjustment.

Method Participants The current study was part of the Kandinsky Longitudinal Study (KLS), a research project aimed at detecting youth at risk for social and emotional problems in The Netherlands (Stoltz, Cillessen, van den Berg, & Gommans, 2016). Participants were adolescents in Grades 7–10 from one high school (first four years of Dutch secondary education). A total of 1,086 adolescents across 41 classrooms were invited to complete a survey at school (51.2% girls, Mage = 14.14, SD = 1.28). The majority of adolescents were born in The Netherlands and had parents who were both born in The Netherlands as well (96.0% and 79.1%, respectively). We omitted 70 cases from all analyses because of absenteeism, lack of consent, or unfinished questionnaires due to time constraints. The remaining 1,016 adolescents indicated how many hours they played video games during a typical week. Those who indicated to never play video games skipped all other items on video game play behavior (n = 311; 30.6% of all students). The final sample consisted of 705 adolescents (33.5% girls, Mage = 14.07, SD = 1.29)1. Data preparation and analyses were done using R version 3.2.2 (R Development Core Team, 2008).

Procedure The KLS research project was formally requested by the school in 2010 to gain insight in the socio-emotional wellbeing of their students. The school signed an agreement in which they claimed responsibility for the parental consent procedure and distributed a letter among the parents describing the purpose and procedure of the longitudinal research project. Parents were requested to respond in order to exclude their child from participation, thus informed parental consent was obtained through the school at the beginning of the school year. Adolescents provided assent at the start of the survey. This procedure was approved by the Institutional Review Board.

In order to investigate differences between total sample and our selection of respondents on adjustment, a series of ANCOVAs were performed. Whether participants played games was used as a dichotomous independent variable, the six indicators of adjustment were dependent variables, and age and gender were included as covariates. Two small significant differences between non-gamers and gamers were found. Compared with adolescents who play games, non-gamers scored higher on their need to belong, F(1, 916) = 4.05, p < .05, η2 = .004, and nongamers scored higher on proactive aggression, F(1, 1012) = 5.47, p < .05, η2 = .005.

Ó 2019 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1), 27–39

The survey was administered at school in a classroom using laptops and lasted 45–60 min. Plenary verbal instructions about the goal of the study and the computerized questionnaire were provided at the start of the assessment session. All adolescent sat at a private desk and talking was prohibited. At least two researchers were present during data collection.

Measures Video Game Play Behavior Several questionnaires were administered in order to capture the five distinct aspects of gaming behavior. The frequency of gaming was assessed using two questions: “During the week/weekend, how many hours do you play video games?” Answers ranged from 0 (= I never play video games) to 5 (= more than 4 hours per day). Frequency was calculated by taking the score for the week question and multiplying it by five, then adding the score for the weekend question multiplied by two. This approach has been used successfully in previous studies (e.g., Brooks et al., 2016; Heim et al., 2007; Yang et al., 2013). To measure the social context in which adolescents play video games, adolescents indicated “In what way do you usually play videogames?” on a 4-point scale (1 = alone, 2 = together with strangers online, 3 = together with acquaintances online, 4 = together with acquaintances offline [for instance, on the couch]). Strangers were defined as “People whom you have never met in real life (only know through the Internet)” and acquaintances as “For instance, friends or family.” The adolescents’ identification as a gamer was measured using the question “To what degree would you describe yourself a gamer?” Answers were rated on a 6-point scale (1 = not at all, 6 = all the way). Adolescents indicated which genres they played based on the multiple-choice question “What kind of video games have you played in the past 12 months?” Answers were binary (0 = did not play this genre, 1 = have played this genre). The 16 different genres included Racing, Puzzle, Sports, Action, Adventure, Platformer, Rhythm, Strategy, Simulation, Fighting, First-Person Shooters, Role-Playing, Survival Horror, MMOGs, Virtual Worlds, and Social Network Games (Lenhart et al., 2008). For each genre, three popular games were given as an example. To reduce the number of genres, an exploratory factor analysis with principal axis factoring was performed on the tetrachoric correlation matrix using the psych package in R (Revelle, 2015). Four factors exceeded Kaiser’s criterion of eigenvalues above 1. A four-factor solution with promax rotation indicated that the Strategy genre failed to load above .3 on any factor. Removal of the Strategy genre created a four-factor structure explaining 49% of the variance, with all primary Journal of Media Psychology (2020), 32(1), 27–39

G. P. Verheijen et al., Video Games and Adolescent Adjustment

loadings at or above .4 and all cross-loadings below .4. The first of the four factors was defined as Competition and included genres that are often fast-paced, where players compete with real or computer-controlled opponents (i.e., Action, First-Person Shooters, Fighting, Survival Horror, Adventure, Sport). The second factor was named Mastery and included genres where players are challenged to overcome obstacles and improve upon themselves (i.e., Simulation, Platformer, Racing, Rhythm, Puzzle). The third factor was called Immersion and included genres that are rich in narrative, set in imaginative worlds (i.e., MMO and Role-Playing games). The fourth factor was named Connection and included genres with an emphasis on connecting players with each other (i.e., Virtual World and Social Network games). Table 1 presents the genres for each factor as well as the standardized factor loadings. Factor scores were used in further analyses. Adolescents’ motivation to engage in gaming behavior was measured using the motivation for video game play scale (Ferguson & Olson, 2013). This questionnaire is based on the self-determination theory (Przybylski et al., 2010) and measures motivation for gaming using four subscales: Fun, Catharsis, Social, and Bored. The scale was translated into Dutch by the first author, then back-translated to English by three independent translators and reviewed by an expert panel. Participants indicated their answers on 16 items using a 4-point Likert scale (1 = strongly disagree, 4 = strongly agree). To evaluate the internal structure of the scale in our sample, a confirmatory factor analysis was performed, w2(98) = 447.23, CFI = 0.86, TLI = 0.83, RMSEA = 0.07, using the lavaan package in R (Rosseel, 2012). Based on this, two items with the highest modification indices were removed (“It helps me relax” and “I like to learn new things”), which substantially improved model fit, w2(71) = 223.87, CFI = 0.92, TLI = 0.89, RMSEA = 0.06. Cronbach’s α values were acceptable for subscales fun (.75) and catharsis (.68), but low for social (.59) and bored (.55). The individual items and their standardized factor loadings can be found in Appendix. The sum scores of each subscale were used in further analyses. Adolescent Adjustment Measures of adjustment included internalizing behaviors, namely, loneliness, self-esteem, need to belong, and fear of negative evaluation. Externalizing behaviors included proactive and reactive aggression. Loneliness was measured with the Loneliness and Aloneness Scale for Children and Adolescents (Marcoen, Goossens, & Caes, 1987). Adolescents rated each item on a 4-point scale (0 = never, 1 = seldom, 2 = sometimes, 4 = often; α = .90). Self-Esteem was measured with the Rosenberg Self-Esteem Scale (Rosenberg, 1965). Adolescents rated each statement on a 4-point scale (1 = strongly Ó 2019 Hogrefe Publishing

G. P. Verheijen et al., Video Games and Adolescent Adjustment

Table 1. Factor loadings based on a principal axis factoring with promax rotation for 15 genres of video games (N = 705) Competition Action

.86

First-Person Shooters

.62

Fighting

.60

Survival Horror

.49

Adventure

.41

Sport

.40

Mastery

Simulation

.70

Platformer

.67

Racing

.62

Rhythm

.51

Puzzle

.43

Immersion

Massive Multiplayer Online

.90

Role-Playing Games

.55

Virtual World

.72

Social Network Games % of variance explained

Connection

.60 16

Note. Factor loadings < .40 are not shown.

disagree, 4 = strongly agree; α = .88). The Need to Belong Scale (Leary, Kelly, Cottrell, & Schreindorfer, 2013) was used to measure the desire for acceptance and belonging. Adolescents rated each item on a 6-point scale (1 = not at all characteristic of me, 6 = extremely characteristic of me; α = .67). Fear of negative evaluation was measured with the Brief Fear of Negative Evaluation Scale (Leary, 1983). Adolescents rated how much they experienced apprehension at the prospect of being evaluated negatively on a 6-point scale (1 = not at all characteristic of me, 6 = extremely characteristic of me; α = .86). For all self-report questionnaires of adjustment, sum scores were used in further analyses. Proactive and reactive aggression were measured separately with peer nominations. Participants received a roster with the names of their classmates and were asked to nominate peers that fitted descriptions based on the Reactive– Proactive Aggression Questionnaire (Dodge & Coie, 1987). For each question, participants could nominate as many peers as they saw fit, but not themselves. The total number of nominations received from classmates, standardized within classrooms, were used in analyses.

Results Descriptive Statistics The descriptive statistics and intercorrelations between study variables (see Table 2) were calculated using the psych package in R (Revelle, 2015). Of the total sample, Ó 2019 Hogrefe Publishing

705 adolescents (69.4%) indicated playing video games for at least 1 hr per week, with an average playing time of 14.12 hr per week. Correlations showed small to moderate positive associations between different aspects of video game play behavior. Thus, while there was overlap between the separate gaming variables, effect sizes indicated that the different aspects of gaming behavior were not redundant. There were several other notable intercorrelations between variables of gaming behavior. Most subscales of motivation for playing video games correlated positively with each other, indicating that adolescents have a variety of reasons for engaging in gaming. Furthermore, Catharsis motivation was linked to other aspects of gaming behavior most often, while the Bored motivation and the Connection genre were mostly independent of other aspects.

Overall Findings A series of hierarchical multiple regression analyses – one for each outcome variable – was conducted to first show the association between frequency of gaming and adolescent adjustment and then the impact of adding other variables of video game play behavior as predictors. A structural equation modeling approach was used in order to allow imputation of missing cases. This was done in the R package lavaan using maximum likelihood estimations and Huber–White robust standard errors (Rosseel, 2012). Gender, age, and frequency were added in Step 1 and the other four aspects of video game play behavior (social context, gamer identification, genre factors, and motivation subscales) were entered in Step 2. There were Journal of Media Psychology (2020), 32(1), 27–39

Journal of Media Psychology (2020), 32(1), 27–39

7.54

705

1.10

2.31

.65***

.10

705

1.62

2.91

.17***

705

0.87

0.00

.44***

.23***

.31***

705

0.94

0.00

.21***

.00

.09

.08

705

1.40

0.00

.38***

.31***

705

0.85

0.00

.13

.12

.08

.10

.05

.09

.17**

.00

.11

694

2.90

12.13

.00

.09

.04

.44***

.56***

.20***

.43***

694

2.76

7.16

.41***

.23***

.18***

.11

.17**

.38***

.05

.27***

694

2.49

6.66

.48***

.51***

.11

.21***

.01

.31***

.49***

.16**

.44***

Note. Neg. Eval. = negative evaluation. *p < .05, **p < .01, ***p < .001. p values were obtained after Bonferroni correction.

14.12

17 Reactive Aggression

16 Proactive Aggression

15 Fear of Neg. Eval.

14 Need to Belong

13 Self-Esteem

12 Loneliness

11 Bored Motivation

10 Social Motivation

9 Catharsis Motivation

8 Fun Motivation

7 Connection Genre

6 Immersion Genre

5 Mastery Genre

4 Competition Genre

3 Gamer Identification

2 Social Context

1 Frequency

Variables

Table 2. Descriptive statistics and correlations of model variables

694

1.64

4.94

.13

.24***

.15*

.04

.02

.01

.03

.07

.10

634

5.35

16.53

.05

.03

.28***

.05

.22***

.10

.14

.04

.02

.17**

.02

642

5.08

32.15

.30***

.51***

616

6.34

37.01

.27***

.02

.10

.07

.02

.04

.00

.11

.15*

.03

.14

.12

.04

.21***

.19***

.10

.05

.09

.07

.13

.12

.13

640

10.22

38.81

.62***

.50***

.45***

.06

.08

.16**

.11

.05

.02

.06

.10

.09

.14

705

1.04

0.05

.03

.10

705

1.05

0.07

.56***

.02

.03

.21***

.04

.06

.21***

.12

.06

.04

.02

.10

.19***

.00

.14*

.04

.08

.02

.04

.02

.13

.03

.00

.14*

.16**

.09

.12

.09

32 G. P. Verheijen et al., Video Games and Adolescent Adjustment

Ó 2019 Hogrefe Publishing

G. P. Verheijen et al., Video Games and Adolescent Adjustment

Table 3. Hierarchical multiple regression model of video game play behavior Loneliness B Step 1 Gender

ΔR2

Self-Esteem B

.02*** .04

ΔR2

Need to Belong β

.10*** .29***

ΔR2

Fear of Negative Evaluation β

.04*** .14*

ΔR2

Proactive Aggression

Reactive Aggression

.08*** .28***

ΔR2 .07***

.25***

.04*** .21***

Age

.02

.03

.00

.09*

.02

.01

Frequency

.02

.04

.07

.00

.01

Step 2

.13***

.10***

Social Context

.10*

.00

Gamer Identification

.05

.02

Competition Genre

.04

.11*

.04** .05

.06***

.03*

.05***

.00

.05

.01

.10

.04

.09

.02

.01

.11*

.01 .03

Mastery Genre

.02

.06

.05

.04

.14***

Immersion Genre

.04

.03

.01

.04

.01

Connection Genre

.12**

.03

.02

.05

.06

.05

Fun Motivation

.11*

.21***

.04

.05

.01

.05

Catharsis Motivation

.31***

.24***

.29***

.04

.21***

Social Motivation

.06

.00

.11*

.10

.07

.13*

Bored Motivation

.02

.09*

.05

.01

.03

.00

Total R2

.15***

.20***

.08***

.14***

ΔR2

.10***

.09***

Note. *p < .05, **p < .01, ***p < .001.

no indications of autocorrelation or multicollinearity. Results are presented in Table 3. Frequency of play did not significantly explain variation on any outcome variable. While a significant proportion of variance was explained for each adjustment outcome in Step 1, this was primarily due to gender. Girls scored higher on loneliness (β = .15, p < .01), need to belong (β = .15, p < .01), and fear of negative evaluation (β = .27, p < .001), and lower on self-esteem (β = .32, p < .001), proactive aggression (β = .28, p < .001), and reactive aggression (β = .17, p < .001) than boys. Step 2 showed that including other aspects of gaming behavior contributed significantly to the regression model. All outcome variables showed a significant change in the proportion of variance explained. This change was largest for loneliness, w2(10, N = 705) = 83.79, p < .001, ΔR2 = .13, and smallest for proactive aggression, w2(10, N = 705) = 19.69, p < .05, ΔR2 = .03. The significant changes in proportion of variance explained in Step 2 confirms our expectation of the added value of considering different aspects of gaming behavior when explaining adolescent adjustment.

Detailed Findings Several aspects of video game play behavior emerged as significant predictors of adolescent adjustment, after controlling for age, gender, and frequency. These include the social context, type of game, and motivation for playing. Ó 2019 Hogrefe Publishing

There was a significant negative association between social context and loneliness. Adolescents who usually play video games in a more social setting also reported less loneliness. No other significant associations between social context and adjustment were found. Furthermore, the degree to which an adolescent identified as a gamer did not predict adolescent adjustment. Four significant associations between the type of game and adjustment were found. Loneliness was positively associated with the Connection genre, while self-esteem was negatively related to the Competition genre. Interestingly, the reverse associations were not significant; Competition was not associated with loneliness nor Connection with self-esteem. This illustrates that different types of video games can uniquely predict different domains of adjustment. Furthermore, proactive aggression was positively associated with the Competition genre but negatively associated with the Mastery genre. This shows that even within the same domain of adolescent adjustment, both positive and negative associations with gaming can exist. The motivation subscales showed several associations with adjustment. Most notable was the Catharsis motivation, which was associated with five of the six adjustment outcomes. The regression coefficients consistently showed that playing games for catharsis was associated with poorer adjustment. Bored motivation also was negatively associated with self-esteem. By contrast, the other two motivation subscales predicted positive adjustment outcomes. Fun motivation was negatively associated with loneliness and Journal of Media Psychology (2020), 32(1), 27–39

positively with self-esteem. Social motivation was negatively associated with need to belong and reactive aggression. This confirms again that gaming behavior has both positive and negative associations with adjustment. We performed additional analyses to investigate whether frequency of play moderated associations of other game aspects (Gentile, 2011). For this, interaction terms between frequency and every other gaming predictor were entered in Step 3 of the hierarchical regression analyses. A significant change from Step 2 to Step 3 was found only for self-esteem, where adding the interaction effects explained an additional 3% of variance, w2(10, N = 705) = 24.73, p < .01. The only significant interaction was found between frequency and Competition genre (β = .13, p < .01). Simple slopes tests comparing low (1 SD below the mean), moderate (mean), and high (1 SD above the mean) levels of frequency showed a significant negative association between the Competition genre and self-esteem for low (b = 1.41, SEb = .35, p < .001) and moderate frequency (b = 0.75, SEb = .26, p < .01), but not for high levels of frequency (b = 0.083, SEb = .33, p = .802). Thus, playing competitive games was associated with lower self-esteem for adolescents with low or moderate frequency of play.

Discussion The aim of the current study was to demonstrate the unique associations between separate aspects of gaming behavior and various measures of adolescent adjustment. We focused on five specific aspects of video game play behavior, namely, frequency of gaming, social context, gamer identification, type of game, and motivation for playing games. Hierarchical regression analyses showed that the social context of gaming, type of game, and motivation for playing were significantly associated with adjustment. However, frequency was not significantly associated with any outcome variable. This goes against displacement theory, which predicts that the amount of hours playing video games will have a direct negative impact on youth (Hofferth, 2010). It also underlines the necessity to look beyond frequency in order to fully understand the impact of gaming on youth. Furthermore, different aspects of video game behavior were found to be uniquely associated with separate measures of adjustment, and even showed opposite associations within the same outcome variable. Both positive and negative associations with adjustment emerged, depending on the aspect of gaming behavior. This emphasizes the necessity to control for the different aspects of video game play behavior when predicting the impact on adjustment, since both deleterious and beneficial associations exist. Journal of Media Psychology (2020), 32(1), 27–39

G. P. Verheijen et al., Video Games and Adolescent Adjustment

Associations Between Video Game Play Behavior and Adolescent Adjustment Earlier studies have found that a high frequency of gaming was linked to lower health outcomes, such as lower general well-being (Przybylski, 2014; Yang et al., 2013), self-esteem (Colwell & Payne, 2000), and sociability (Griffiths, 2010; Padilla-Walker et al., 2010). However, frequency of gaming was not significantly associated with any measure of adolescent adjustment in our sample. These results corroborate findings from other studies that have also reported the lack of a statistically significant relation between frequency and these outcomes (Colwell & Kato, 2003; Ferguson, 2015; Willoughby, 2008). There also was no evidence that frequency exacerbates the effect of other aspects of video game play behavior (Gentile, 2011). Moderation of frequency was only found for self-esteem, and simple slopes tests indicated that the negative association between the Competition genre and self-esteem was larger for participants with lower levels of frequency. When looking at aspects beyond the frequency of gaming, findings are mixed. Playing games in a more social context was associated with a lower score on loneliness. This is in line with suggestions that adolescents often use video games as a focus for hanging out (Olson, Kutner, & Warner, 2008). Yet this result should be interpreted with caution as the effect size was small, and no associations were found between social context and adolescents’ need to belong or fear of negative behavior. Similarly, gamer identification was not associated with adolescent adjustment at all. While negative stereotypes of gamers being socially incompetent exist (Kowert et al., 2012), no relation with this stigmatization was found in our sample. Regarding type of games, the genre factors Competition, Mastery, Immersion, and Connection showed four significant associations with adolescent adjustment. First, playing the Connection genre was associated with increased loneliness. Previous research suggests that this association might be driven by adolescents who feel lonely and seek out online games for social compensation (Kowert et al., 2015), but causal directionality is unclear in our sample. Next, a negative association between the Competition genre and self-esteem was found. This might have emerged as adolescents with low self-esteem seek out competitive games in order to enhance their self-esteem by way of social comparisons (Vorderer, Hartmann, & Klimmt, 2003). The interaction effect suggested that this is especially true for adolescents with low frequency of gaming. However, given the small change in variance explained by the interaction, we recommend future research to replicate this finding. Lastly, proactive aggression was distinctively associated with both the Mastery genre and the Competition genre. Ó 2019 Hogrefe Publishing

G. P. Verheijen et al., Video Games and Adolescent Adjustment

The Mastery genre showed a negative association with proactive aggression. Proactive aggression is characterized by using aggressive behavior to reach a goal (Dodge & Coie, 1987), something that is frequently required in video games. Yet the Mastery genre is defined by genres that focus on single-player mode of play, where players overcome obstacles and improve upon themselves without necessarily defeating opponents. Perhaps the negative association emerged because adolescents who refrain from proactive aggression in daily life also prefer genres where proactive aggression is minimal. By contrast, a positive association was found between the Competition genre and proactive aggression. This association was hypothesized, as Competitive play has been linked to aggressive behavior before (Adachi & Willoughby, 2016). The games in the Competition genre tended to be high in violent content as well, which has also been linked to aggression (Anderson et al., 2010), although some researchers argue that the impact of violent gaming on youth is minimal (Ferguson, 2015). Whether the association between violent games and aggression is the result of methodological issues (Hilgard, Engelhardt, & Rouder, 2017) or indicative of a true effect (Kepes, Bushman, & Anderson, 2017) remains disputed. In the current study, it is important to note that we cannot distinguish whether either competition or violent content was the driving force behind the association between competitive genres and aggression. The four subscales of motivation for playing games showed distinct associations with adjustment as well. Most notably, Catharsis motivation was repeatedly associated with poorer adjustment. Previous research has argued that playing games for cathartic reasons might be beneficial, as it can be used for self-regulatory purposes such as improving one’s mood (Russoniello, O’Brien, & Parks, 2009). On the other hand, the suggestion that cathartic video games can help vent anger and reduce aggression has not received much support in empirical research (Gentile, 2013) and using video games for escapism in order to cope with life problems may lead to more negative outcomes in the long run (Kardefelt-Winther, 2014). Our results point to a negative association between catharsis and adjustment, but it is important to note that causality cannot be inferred. Regardless, based on the number and strength of the relationships between Catharsis motivation and adjustment, the impact of cathartic video games seems a particular worthwhile avenue for future research.

self-identification as a gamer, which were assessed using only a single item. Similarly, the measure for gaming frequency was cut off at 4 hr or more per day. The genre that participants played was also only measured using dichotomous questions, while a Likert scale would have been better able to distinguish between frequent and one-time players of a genre. However, the aim of this study was not to be exhaustive in measurement but rather to provide initial evidence for the complexity of gaming behavior. We encourage future gaming surveys to always include multiple aspects of game play behavior or zoom in on specific associations found in the current study. Furthermore, our cross-sectional research design makes it impossible to draw directional conclusions. Whether gaming behavior influences adjustment and well-being, or certain levels of adjustment and well-being determine video game play behavior, remains unanswered in this study. Longitudinal studies are needed to clarify the gaming patterns and developmental trajectories of adolescents over time. A strength of this study is that a large, non-convenience sample of adolescents was investigated. We did not, however, investigate pathological play in our study. While problematic play is associated with a high frequency, it is also characterized by obsessiveness and loss of control over play time (Lemmens, Valkenburg, & Peter, 2009) and related to lower well-being (Przybylski et al., 2009). Thus, our findings might not be applicable to adolescents who engage in problematic play, especially concerning the positive associations between gaming and adjustment. Finally, it is important to note that most effect sizes found in this study were small. Caution is warranted when drawing conclusions of practical significance. It has been argued that most gaming effects are small in size, and actually trivial on a societal level (Ferguson, 2015). The degree to which gaming behavior is associated with societal outcomes remains an open question that should not be answered by scholars alone, but also by policy makers and the general public. Still, while the significant effect sizes were small, they remain significant after controlling for 10 predictors and two covariates, speaking to the robustness of effects. Furthermore, some aspects of video game play behavior were more influential than others. Particularly the Catharsis motivation showed larger, consistent associations with adjustment, and should be explored in further research.

Conclusion Strengths and Limitations Gaming behavior encompasses more than the aspects used in the current study, and our gaming questionnaire was shortened to prevent participants’ fatigue. This limitation is evident for measurements of the social context and Ó 2019 Hogrefe Publishing

This study took a nuanced approach toward video game play behavior to help understand the complex relation between gaming and adolescent adjustment. We separated video game play behavior into five distinct aspects, namely, frequency of play, social context, adolescents’ identification Journal of Media Psychology (2020), 32(1), 27–39

as a gamer, type of game, and motivation for playing. The associations between gaming behavior and adjustment were illustrated using a large normative sample of adolescents. The results underline the value of treating video game play behavior as a multidimensional construct and to consider the distinct and unique associations with adolescent adjustment. Ideally, future research should disaggregate aspects of video game play behavior and continue to investigate the multidimensionality of the impact of video games on youth.

Research Transparency Statement The authors are willing to share their data, analytics methods, and study materials with other researchers. The material will be available upon request. The authors have not preregistered this research.

References Aarsand, P. (2012). The ordinary player: Teenagers talk about digital games. Journal of Youth Studies, 15, 961–977. https:// doi.org/10.1080/13676261.2012.685150 Adachi, P. J. C., & Willoughby, T. (2016). The longitudinal association between competitive video game play and aggression among adolescents and young adults. Child Development, 87, 1877–1892. https://doi.org/10.1111/cdev.12556 Allahverdipour, H., Bazargan, M., Farhadinasab, A., & Moeini, B. (2010). Correlates of video games playing among adolescents in an Islamic country. BMC Public Health, 10, 1–7. https://doi.org/ 10.1186/1471-2458-10-286 Anderson, C. A., Shibuya, A., Ihori, N., Swing, E. L., Bushman, B. J., Sakamoto, A., . . . Saleem, M. (2010). Violent video game effects on aggression, empathy, and prosocial behavior in Eastern and Western countries: A meta-analytic review. Psychological Bulletin, 136, 151–173. https://doi.org/10.1037/a0018251 Brooks, F. M., Chester, K. L., Smeeton, N. C., & Spencer, N. H. (2016). Video gaming in adolescence: Factors associated with leisure time use. Journal of Youth Studies, 19, 36–54. https:// doi.org/10.1080/13676261.2015.1048200 Casiano, H., Kinley, D. J., Katz, L. Y., Chartier, M. J., & Sareen, J. (2012). Media use and health outcomes in adolescents: Findings from a nationally representative survey. Journal of the Canadian Academy of Child and Adolescent Psychiatry, 21, 296–301. Colwell, J., & Kato, M. (2003). Investigation of the relationship between social isolation, self-esteem, aggression and computer game play in Japanese adolescents. Asian Journal of Social Psychology, 6, 149–158. https://doi.org/10.1111/1467839X.t01-1-00017 Colwell, J., & Payne, J. (2000). Negative correlates of computer game play in adolescents. British Journal of Psychology, 91, 295–310. https://doi.org/10.1348/000712600161844 De Grove, F., Courtois, C., & van Looy, J. (2015). How to be a gamer! Exploring personal and social indicators of gamer identity. Journal of Computer-Mediated Communication, 20, 346–361. https://doi.org/10.1111/jcc4.12114 Dodge, K. A., & Coie, J. D. (1987). Social-information-processing factors in reactive and proactive aggression in children’s peer groups. Journal of Personality and Social Psychology, 53, 1146–1158. https://doi.org/10.1037/0022-3514.53.6.1146

Journal of Media Psychology (2020), 32(1), 27–39

G. P. Verheijen et al., Video Games and Adolescent Adjustment

Durkin, K., & Barber, B. (2002). Not so doomed: Computer game play and positive adolescent development. Journal of Applied Developmental Psychology, 23, 373–392. https://doi.org/ 10.1016/S0193-3973(02)00124-7 Ferguson, C. J. (2015). Do Angry Birds make for angry children? A meta-analysis of video game influences on children’s and adolescents’ aggression, mental health, prosocial behavior, and academic performance. Perspectives on Psychological Science, 10, 646–666. https://doi.org/10.1177/1745691615592234 Ferguson, C. J., & Olson, C. K. (2013). Friends, fun, frustration and fantasy: Child motivations for video game play. Motivation and Emotion, 37, 154–164. https://doi.org/10.1007/s11031-012-9284-7 Gentile, D. A. (2011). The multiple dimensions of video game effects. Child Development Perspectives, 5, 75–81. https://doi. org/10.1111/j.1750-8606.2011.00159.x Gentile, D. A. (2013). Catharsis and media violence: A conceptual analysis. Societies, 3, 491–510. https://doi.org/10.3390/ soc3040491 Granic, I., Lobel, A., & Engels, R. C. M. E. (2014). The benefits of playing video games. American Psychologist, 69, 66–78. https:// doi.org/10.1037/a0034857 Griffiths, M. D. (2010). Computer game playing and social skills: A pilot study. Aloma: Revista de Psicologia, Ciències de l’Educació i de l’Esport, 27, 301–310. Heim, J., Brandtzæg, P. B., Kaare, B. H., Endestad, T., & Torgersen, L. (2007). Children’s usage of media technologies and psychosocial factors. New Media & Society, 9, 425–454. https://doi.org/10.1177/1461444807076971 Hilgard, J., Engelhardt, C. R., & Rouder, J. N. (2017). Overstated evidence for short-term effects of violent games on affect and behavior: A reanalysis of Anderson et al. (2010). Psychological Bulletin, 143, 757–774. https://doi.org/10.1037/bul0000074 Hofferth, S. L. (2010). Home media and children’s achievement and behavior. Child Development, 81, 1598–1619. https://doi. org/10.1111/j.1467-8624.2010.01494.x Johnson, D. W. P., & Sweetser, P. (2013, September). The peoplegame-play model for understanding videogames’ impact on wellbeing. Paper presented at the 5th International IEEE Games Innovation Conference (IGIC 2013), Vancouver, Canada Kardefelt-Winther, D. (2014). The moderating role of psychosocial well-being on the relationship between escapism and excessive online gaming. Computers in Human Behavior, 38, 68–74. https://doi.org/10.1016/j.chb.2014.05.020 Kepes, S., Bushman, B. J., & Anderson, C. A. (2017). Violent video game effects remain a societal concern: Reply to Hilgard, Engelhardt, and Rouder (2017). Psychological Bulletin, 143, 775–782. https://doi.org/10.1037/bul0000112 Kowert, R., Domahidi, E., Festl, R., & Quandt, T. (2014). Social gaming, lonely life? The impact of digital game play on adolescents’ social circles. Computers in Human Behavior, 36, 385–390. https://doi.org/10.1016/j.chb.2014.04.003 Kowert, R., Griffiths, M. D., & Oldmeadow, J. A. (2012). Geek or chic? Emerging stereotypes of online gamers. Bulletin of Science, Technology & Society, 32, 471–479. https://doi.org/ 10.1177/0270467612469078 Kowert, R., Vogelgesang, J., Festl, R., & Quandt, T. (2015). Psychosocial causes and consequences of online video game play. Computers in Human Behavior, 45, 51–58. https://doi.org/ 10.1016/j.chb.2014.11.074 Kühn, S., Kugler, D. T., Schmalen, K., Weichenberger, M., Witt, C., & Gallinat, J. (2018). Does playing violent video games cause aggression? A longitudinal intervention study. Molecular Psychiatry. https://doi.org/10.1038/s41380–018-0031–7 Leary, M. R. (1983). A brief version of the Fear of Negative Evaluation Scale. Personality and Social Psychology Bulletin, 9, 371–375. https://doi.org/10.1177/0146167283093007

Ó 2019 Hogrefe Publishing

G. P. Verheijen et al., Video Games and Adolescent Adjustment

Leary, M. R., Kelly, K. M., Cottrell, C. A., & Schreindorfer, L. S. (2013). Construct validity of the need to belong scale: Mapping the nomological network. Journal of Personality Assessment, 95, 610–624. https://doi.org/10.1080/00223891.2013.819511 Lemmens, J. S., Valkenburg, P. M., & Peter, J. (2009). Development and validation of a game addiction scale for adolescents. Media Psychology, 12, 77–95. https://doi.org/10.1080/ 15213260802669458 Lenhart, A., Kahne, J., Middaugh, E., Macgill, A. R., Evans, C., & Vitak, J. (2008). Teens, video games, and civics: Teens’ gaming experiences are diverse and include significant social interaction and civic engagement. Pew Internet & American Life Project. Retrieved from http://www.pewinternet.org/Reports/2008/ Teens-Video-Games-and-Civics.aspx Lenhart, A., Smith, A., Anderson, M., Duggan, M., & Perrin, A. (2015). Teens, technology and friendships. Pew Internet and American Life Project. Retrieved from http://www.pewinternet. org/2015/08/06/teens-technology-and-friendships/ Maras, D., Flament, M. F., Murray, M., Buchholz, A., Henderson, K. A., Obeid, N., & Goldfield, G. S. (2015). Screen time is associated with depression and anxiety in Canadian youth. Preventive Medicine, 73, 133–138. https://doi.org/10.1016/j. ypmed.2015.01.029 Marcoen, A., Goossens, L., & Caes, P. (1987). Loneliness in prethrough late adolescence: Exploring the contributions of a multidimensional approach. Journal of Youth and Adolescence, 16, 561–577. https://doi.org/10.1007/BF02138821 Markey, P. M. (2015). Finding the middle ground in violent video game research: Lessons from Ferguson (2015). Perspectives on Psychological Science, 10, 667–670. https://doi.org/10.1177/ 1745691615592236 Merritt, A., LaQuea, R., Cromwell, R., & Ferguson, C. J. (2016). Media managing mood: A look at the possible effects of violent media on affect. Child & Youth Care Forum, 45, 241–258. https://doi.org/10.1007/s10566-015-9328-8 Olson, C. K., Kutner, L. A., & Warner, D. E. (2008). The role of violent video game content in adolescent development: Boys’ perspectives. Journal of Adolescent Research, 23, 55–75. https://doi.org/10.1177/0743558407310713 Padilla-Walker, L. M., Nelson, L. J., Carroll, J. S., & Jensen, A. C. (2010). More than a just a game: Video game and internet use during emerging adulthood. Journal of Youth and Adolescence, 39, 103–113. https://doi.org/10.1007/s10964-008-9390-8 Primack, B. A., Swanier, B., Georgiopoulos, A. M., Land, S. R., & Fine, M. J. (2009). Association between media use in adolescence and depression in young adulthood: A longitudinal study. Archives of General Psychiatry, 66, 181–188. https://doi.org/ 10.1001/archgenpsychiatry.2008.532 Przybylski, A. K. (2014). Electronic gaming and psychosocial adjustment. Pediatrics, 134, E716–E722. https://doi.org/ 10.1542/peds.2013-4021 Przybylski, A. K., Rigby, C. S., & Ryan, R. M. (2010). A motivational model of video game engagement. Review of General Psychology, 14, 154–166. https://doi.org/10.1037/a0019440 Przybylski, A. K., Weinstein, N., Ryan, R. M., & Rigby, C. S. (2009). Having to versus wanting to play: Background and consequences of harmonious versus obsessive engagement in video games. CyberPsychology & Behavior, 12, 485–492. https://doi. org/10.1089/cpb.2009.0083 Ravaja, N., Saari, T., Turpeinen, M., Laarni, J., Salminen, M., & Kivikangas, M. (2006). Spatial presence and emotions during video game playing: Does it matter with whom you play? Presence: Teleoperators and Virtual Environments, 15, 381–392. https://doi.org/10.1162/pres.15.4.381 Revelle, W. (2015). psych: Procedures for personality and psychological research (R package version 1.5.4). Evanston, IL:

Ó 2019 Hogrefe Publishing

Northwestern University. Retrieved from http://CRAN.R-project. org/package=psych Romer, D., Bagdasarov, Z., & More, E. (2013). Older versus newer media and the well-being of United States youth: Results from a national longitudinal panel. Journal of Adolescent Health, 52, 613–619. https://doi.org/10.1016/j.jadohealth.2012.11.012 Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48, 1–36. https://doi. org/10.18637/jss.v048.i02 Russoniello, C., O’Brien, K., & Parks, J. M. (2009). The effectiveness of casual video games in improving mood and decreasing stress. Journal of Cyber Therapy and Rehabilitation, 2, 53–66. Shen, C., & Williams, D. (2011). Unpacking time online: Connecting internet and massively multiplayer online game use with psychosocial well-being. Communication Research, 38, 123–149. https://doi.org/10.1177/0093650210377196 Stoltz, S., Cillessen, A. H., van den Berg, Y. H., & Gommans, R. (2016). Popularity differentially predicts reactive and proactive aggression in early adolescence. Aggressive Behavior, 42, 29–40. https://doi.org/10.1002/ab.21603 R Development Core Team. (2008). R: A language and environment for statistical computing. R Foundation for Statistical Computing: Vienna, Austria. Retrieved from http://www.R-project.org van Dorsselaer, S., Tuithof, M., Verdurmen, J., Spit, M., van Laar, M., & Monshouwer, K. (2016). Jeugd en riskant gedrag 2015 [Youth and Risky Behavior 2015]. Utrecht, The Netherlands: Trimbos Instituut. Vella, K., Johnson, D., & Hides, L. (2013, October 2–4). Positively playful: When videogames lead to player wellbeing. In ACM. (Eds.), Gamification ’13 – Proceedings of the First International Conference on Gameful Design, Research, and Applications (pp. 99–102). New York, NY: ACM. Vorderer, P., Hartmann, T., & Klimmt, C. (2003, May). Explaining the enjoyment of playing video games: The role of competition. Paper presented at the Proceedings of the Second International Conference on Entertainment Computing, Pittsburgh, PA. Wack, E., & Tantleff-Dunn, S. (2009). Relationships between electronic game play, obesity, and psychosocial functioning in young men. CyberPsychology & Behavior, 12, 241–244. https:// doi.org/10.1089/cpb.2008.0151 Wang, S. S. (2014). Farming not alone: Farmville play and the implications on social capital. Social Networking, 3, 230–239. https://doi.org/10.4236/sn.2014.35028 Willoughby, T. (2008). A short-term longitudinal study of internet and computer game use by adolescent boys and girls: Prevalence, frequency of use, and psychosocial predictors. Developmental Psychology, 44, 195–204. https://doi.org/10.1037/ 0012-1649.44.1.195 Witt, E. A., Massman, A. J., & Jackson, L. A. (2011). Trends in youth’s videogame playing, overall computer use, and communication technology use: The impact of self-esteem and the Big Five personality factors. Computers in Human Behavior, 27, 763–769. https://doi.org/10.1016/j.chb.2010.10.025 Yang, F., Helgason, A. R., Sigfusdottir, I. D., & Kristjansson, A. L. (2013). Electronic screen use and mental well-being of 10–12year-old children. The European Journal of Public Health, 23, 492–498. https://doi.org/10.1093/eurpub/cks102

History Received January 15, 2018 Revision received September 6, 2018 Accepted September 28, 2018 Published online March 8, 2019

Journal of Media Psychology (2020), 32(1), 27–39

G. P. Verheijen et al., Video Games and Adolescent Adjustment

Acknowledgments We wish to thank the respondents, teachers, and school administrators who made this research possible. We are grateful to all students who helped during data collection and thank Rob Gommans and Giovanni ten Brink for data processing. Conflict of Interest The authors declare that they have no conflict of interest. Funding This research was supported by the Behavioural Science Institute at Radboud University. ORCID Geert P. Verheijen https://orcid.org/0000-0003-1254-4067 Geert P. Verheijen Behavioural Science Institute Radboud University Montessorilaan 3 6525 HR Nijmegen The Netherlands g.verheijen@bsi.ru.nl

Geert P. Verheijen is a PhD candidate in the Behavioural Science Institute at Radboud University in Nijmegen, The Netherlands. In his research project, he examines the different ways in which average, day-to-day video game play behavior might influence the social behavior and peer relationships of adolescents.

Sabine E. M. J. Stoltz is Assistant Professor of Development Psychology in the Behavioural Science Institute at Radboud University in Nijmegen, The Netherlands. Her research involves the causes and consequences of externalizing and internalizing behavior of youth. She is involved in interventions to prevent bullying behavior at Dutch schools.

Yvonne H. M. van den Berg is Assistant Professor in the Behavioural Science Institute at Radboud University in Nijmegen, The Netherlands. Her research involves the development of social relationships from childhood to adulthood, with a particular interest in the importance of peer relationships and classroom social dynamics.

Antonius H. N. Cillessen is Professor of Developmental Psychology and Director of the Behavioural Science Institute at Radboud University in Nijmegen, The Netherlands. He studies social status and social behavior of children and adolescents and the methodology of developmental psychology research. His publications address peer-group relations, aggression and antisocial behavior, social cognition, and research methods.

William J. Burk is Assistant Professor in the Behavioural Science Institute at Radboud University in Nijmegen, The Netherlands. His research interests broadly concern the impact of close relationships with parents and peers on the psychosocial adjustment of children and adolescents, and he specializes in methodological techniques capable of integrating individual, dyadic, and group-level processes.

Journal of Media Psychology (2020), 32(1), 27–39

Ó 2019 Hogrefe Publishing

G. P. Verheijen et al., Video Games and Adolescent Adjustment

Appendix Table A1. Factor loadings based on a confirmatory factor analysis for 14 items from the child motivation for video game play scale (N = 694) Fun I like the challenge of figuring the game out

.74

I like to compete with other people and win

.71

It’s exciting

.69

It’s just fun

.52

Catharsis

It helps me forget my problems

.74

It helps me get my anger out

.72

It helps me feel less lonely

.58

I like to create my own world

.44

Social

I like to teach other kids how to play

.61

It helps me make new friends

.60

I like to mod games

.45

My friends like to play

.43

There is nothing else to do

.65

It’s something to do when I’m bored Cronbach’s α

Bored

.60 .75

.68

.59

.55

Note. The following items were removed in order to improve the model fit: “It helps me relax” and “I like to learn new things.”

Ó 2019 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1), 27–39

An inspiring volume on the prevention of suicide and suicidal behaviour Diego De Leo / Vita Poštuvan (Editors)

Reducing the Toll of Suicide

Resources for Communities, Groups, and Individuals

New

2020, x / 222 pp. US $36.80 / € 29.95 ISBN 978-0-88937-569-7 Also available as eBook This is the second book based on the Intuition, Imagination and Innovation – TRIPLE i in Suicidology international conferences, which are organised annually by the Slovene Center for Suicide Research in memory of the late Prof. Andrej Marušic with the aim of promoting intuition, imagination, and innovation in the research and prevention of suicide and suicidal behaviour. • Packed with the latest international research • New ideas for interventions – for individuals, groups, and communities • Presents models for understanding suicide • Explores ethical issues

www.hogrefe.com

In five parts, the internationally renowned team of authors summarises the research looking at: • Understanding individuals (assessing risk in older adults and psychotherapy with suicidal patients), • Understanding the groups at risk of suicide (including youth, people in prison, men, and people with mood disorders), • Understanding the role of community (including the Papageno effect, technology-based and collaborative approaches to prevention, as well as bereavement), • Models of understanding suicide (including the integrated motivational-volitional model of suicidal behaviour and the hot air balloon model of risk factors for suicide), and • Understanding the unique ethical and methodological issues associated with research in this field.

Evidence-based diagnosis and treatment for ADHD in children and adolescents “This book could change the way that you treat young people with ADHD.” Steven W. Evans, PhD, Director of Center for Intervention Research in Schools, Professor of Psychology, Ohio University, Athens, OH

Brian P. Daly / Aimee K. Hildenbrand / Ronald T. Brown

ADHD in Children and Adolescents Advances in Psychotherapy – Evidence-Based Practice, vol. 33 2016, viii + 90 pp. US $29.80 / € 24.95 ISBN 978-0-88937-412-6 Also available as eBook Attention-Deficit / Hyperactivity Disorder (ADHD) is a common childhood disorder that can have serious consequences for academic, emotional, social, and occupational functioning. When properly identified and diagnosed, however, there are many interventions for the disorder that have established benefits. This volume is both a compact “how to” reference, for use by professional clinicians in their daily work, and an ideal educational reference for practice-oriented students. The most important feature of this volume is

www.hogrefe.com

that it is practical and “reader friendly”. It has a similar structure to others in the series, and is a compact and easy to follow guide covering all aspects of practice that are relevant in real life in the assessment and management of ADHD across the life span. Tables, relevant case studies, and marginal notes assist orientation, while suggestions for further reading, support groups, and educational organizations are provided for individuals and professionals. A companion volume on ADHD in Adults is also available in the series.

Research Report

Exploring the Effects of Personal Information in Television News A Cognitive Approach Miguel Barreda-Ángeles1 , Julia Virgós2, Maddalena Fedele3, and Alexandre Pereda-Baños1 Eurecat – Technology Centre of Catalonia, Barcelona, Spain

1 2

University of Barcelona, Barcelona, Spain

School of Communication and International Relations, Ramón Llull University, Barcelona, Spain Abstract: We report the results of an experiment in which participants (N = 41) watched a number of television news clips containing either personal or professional stories about celebrities, while measures of information encoding, storage, and retrieval were collected together with psychophysiological measures of autonomous nervous system activity. The results show that the presence of personal information elicits attentional resources allocation and improves memory for contents, suggesting that personal information benefits from a deeper cognitive processing compared with professional information. These findings provide preliminary evidence on the cognitive effects of personal information that encourages further research on several dimensions of media use such as fandom or gossip media from a cognitive point of view. Keywords: personal information, attention, memory, psychophysiology, gossip

Information about the personal affairs of famous people has an overwhelming presence in the media landscape – usually in the form of celebrity gossip – and audiences show a marked interest in the lives of people they will probably never meet and in events that have no bearing on their daily matters (Turner, 2004). However, and rather surprisingly, research on media psychology has seldom analyzed the effects of personal information in terms of its effects on cognitive processing. In the present report, we argue that information on personal affairs of famous people is, due to its motivational relevance, preferentially processed by our cognitive system. We report the results of an experiment providing preliminary support of this hypothesis in the context of celebrity news and discuss possible next steps for further research in this topic.

Personal Information, Motivational Relevance, and Information Processing Although everyone has a notion of what personal information is, the literature in the field has yet to provide a formal definition of personal information. In general terms, personal information is information about individuals that is not about Journal of Media Psychology (2020), 32(1), 40–45 https://doi.org/10.1027/1864-1105/a000254

their professional activities and includes aspects of their social relationships or private circumstances (e.g., health or socioeconomic status) that have an impact on social life. The exchange of personal information is a defining characteristic of gossip (Foster, 2004). Various authors have considered gossip as a mechanism with an evolutionary basis that has the function of facilitating the exchange of information that is important for group bonding (Dunbar, 2004) or social learning (Baumeister, Zhang, & Vohs, 2004). Therefore, if gossip has such an adaptive function, it seems plausible that our cognitive systems would prime the processing of information serving that function (i.e., personal information), in a similar way in which it favors the processing of other kinds of motivationally relevant information (Lang, 2006). Although one may consider some media contents as professional gossip a quick glance at gossip-related media promptly shows that the more personal the information, the higher the interest it elicits. This goes in line with theories of interpersonal communication such as social penetration theory (Carpenter & Greene, 2016) or communication privacy management theory (Petronio & Durham, 2008), which assign a higher intrinsic informational value to information about people as it gets more personal. And it is precisely this higher informational value that leads to a possible operationalization of this concept, its motivational relevance. Motivational relevance refers to the extent to which information contained in a stimulus represents an opportunity Ó 2019 Hogrefe Publishing

M. Barreda-Ángeles et al., Cognitive Processing of Personal News

or a threat (Lang, 2006), and it has effects on the way information is processed and remembered by the audience. Such effects have been examined primarily within the theoretical framework of the limited capacity model for motivated mediated message processing (LC4MP; Lang, 2006). Within this framework, motivationally relevant information is often operationalized through measures of processing such as performance on the secondary reaction time task (STRT), or through self-reported or physiological measures of sympathetic activation, often referred to as arousal. Arousing contents require more resources to be processed, resulting in slower reaction times in a Secodary Task Reaction Time (STRT) test, but also improve information encoding, storage, and retrieval, which is evidenced in better results in recognition, cued recall, and free recall tests, respectively (e.g., Lang, Park, Sanders-Jackson, Wilson, & Wang, 2007). If personal information, as the object of gossip, represents an opportunity for learning social norms (Baumeister et al., 2004) or controlling free-riders (Dunbar, 2004), it can be considered as motivationally relevant, and then news containing personal information will produce (Hypothesis 1, H1) slower STRTs, (Hypothesis 2, H2) better recognition of story information, along with better (Hypothesis 3, H3) cued recall and (Hypothesis 4, H4) free recall of story information, compared with news containing only professional information. Motivational relevance affects the activity of both the sympathetic and the parasympathetic branches of the autonomous nervous system (Bradley, Codispoti, Cuthbert, & Lang, 2001). Sympathetic activation is associated with increases in electrodermal activity (EDA) levels, while parasympathetic activation has been related to phasic decreases in heart rate (HR) and increases in heart rate variability (HRV; Laborde, Mosley, & Thayer, 2017). Therefore, we also expect that when compared with professional news, personal news will be (Hypothesis 5, H5) positively associated with EDA level, (Hypothesis 6, H6) negatively associated with HR, and (Hypothesis 7, H7) positively associated with HRV.

international celebrities, including four Hollywood actors, three singers, one international model, one football player, and one cinema director. For each celebrity, one story on personal affairs and one story on professional affairs was included. Personal stories were about celebrities getting married (two stories), their bad relationships with relatives or ex-partners (five stories), the birth of a child (one story), the death of one celebrity’s son (one story), and one celebrity admitting he is gay (one story). Professional stories were about the release of a new album or movie (three stories), a new concert (two stories), a movie shooting (one story), a new advertising campaign (one story), or the celebrity receiving an award or acknowledgment (three stories). The formal structure of all of them included the presentation of an edited video with a voice-over narration. There was not a significant difference in their mean duration between personal (M = 65.6; SD = 21.72) and professional (M = 74.5; SD = 16.53) stories, t(18) = 1.03, p = .31, d = 0.46. Regarding the pace of camera changes, measured in camera changes per second, there was no significant difference between personal (M = 0.33; SD = 0.15) and professional (M = 0.46; SD = 0.29) stories, t(18) = 1.34, p = .20, d = 0.60.

Stimuli Pretest

We conducted an experiment in which we asked participants to attentively watch a number of television news stories about celebrities, including stories about personal affairs and stories about professional affairs.

We asked 48 participants (26 women) aged between 19 and 30 years (M = 20.23; SD = 1.99) who watched the 20 stories to rate their agreement with the sentence, “This piece of news is about personal information” on a 7-point Likerttype scale, ranging from 1 (= not at all) to 7 (= totally), in order to obtain a measure of the level of personal information of each story. Participants in the pretest also rated how each story made them feel in terms of emotional arousal and valence using the 5-point Self-Assessment Manikin (SAM) Scale (with a bipolar scale for valence). The mean scores for personal information, arousal, and valence for each story in the sample were calculated, and, in order to compare the two groups of stories, two-sample t tests were applied. The level of personal information was significantly higher in personal stories (M = 5.92; SD = 0.22) than in professional stories (M = 3.19; SD = 0.71), t(18) = 11.61, p < .001, d = 5.19, but there were not significant differences in terms of arousal (M = 2.27; SD = 0.72 for personal stories; M = 2.13; SD = 0.47 for professional stories), t(18) = 0.53, p = .60, d = 0.24. Personal stories (M = 2.59; SD = 0.59) were seen as more negative than professional stories (M = 3.18; SD = 0.47), t(18) = 2.49, p = 0.02, d = 1.11.

Materials

Dependent Variables

We selected 20 stories with a duration of between 27 and 95 s (M = 70.1; SD = 19.37) from actual television programs, all of them about very famous and well-known national or

EDA Participants’ EDA was collected during the viewing of the materials, by placing two electrodes in the middle phalanx

Method

Ó 2019 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1), 40–45

M. Barreda-Ángeles et al., Cognitive Processing of Personal News

of the index and middle fingers of the participant’s nondominant hand.

that they were going to take part in an experiment about the perception of news. They received instructions for the secondary task, and watched a training news reel to habituate to it. The videos were presented on a computer screen in random order using the Psychopy software (Peirce, 2007), which also collected the STRTs. A 5-s-long gray screen was used as a separator between stories. Psychophysiological measures were collected using the Biopac MP-150 system and the Acqknowledge 4.2 (Biopac Systems, Inc.) software (1,000 Hz sampling rate). After the viewing, participants completed the recognition task and, approximately 48 hr later, participants were phoned to answer the free and cued recall questionnaires.

HR and HRV An electrocardiogram (ECG) of the participants was obtained during the viewing sessions, from which the HR and HRV were calculated. Although other alternatives are available (Keene, Clayton, Berke, Loof, & Bolls, 2017), since we focus on tonic effects, HR was measured in beats per minute (bpm). The unit for HRV measurement was the root mean square of successive differences (RMSSD; Laborde et al., 2017). In order to collect the ECG, two electrodes were placed on the non-dominant forearm (one on the wrist and one near the elbow) and one on the wrist of the dominant hand. STRT Seven tones were inserted in pseudo-random positions in each story, except in the case of the two shortest news clips, in which only five tones were inserted. They consisted of a sinusoid waveform with a frequency of 1,000 Hz and a duration of 0.2 s. Their positions varied for each story and each participant, but in all cases they were placed at least 2 s after the beginning of the video and there was a separation of at least 3 s between tones. Recognition A forced choice (Yes/No) auditory recognition task included three 2-s excerpts of the audio of each story, embedded in a similar number of segments of other similar stories about other celebrities not included in the experiment. Participants were asked to detect whether or not each segment had appeared in the viewed materials. Cued and Free Recall Approximately 48 hr after the experiment, a researcher telephoned the participants in order to conduct the cued and free recall tests. The free recall test involved asking the participant to mention the topics of the stories that he or she could recall. Immediately after this, participants carried out the cued recall test, in which they answered two questions about specific details of each story.

Participants and Procedure The participants of the experiment comprised 41 volunteers1 (20 women) aged between 18 and 42 years (M = 25.17; SD = 5.95), who received a gift card as compensation. Each participant completed the experiment individually, seated in a comfortable chair in an isolated room. After signing an informed consent form, participants were told 1

Data Preprocessing Owing to technical issues, the complete physiological recordings of two participants, and the recordings for three stories for one participant, were lost. The ECG for six stories belonging to four participants presented a noisy signal in which heart beats were not properly identifiable, and therefore they were discarded. Thus, the final sample consists of EDA recordings for 777 stories, and ECG recordings for 771 stories, from 39 participants. EDA and HR (obtained from the ECG) were downsampled to 1 Hz. RMSSD was calculated for each second of the story using a 10-s sliding window with 90% overlap (Laborde et al., 2017), over the ECG signal (thus leaving aside the first and last 5 s of each story), in order to obtain a temporal series of HRV. Thus, the data submitted to the analysis are a temporal series, with a 1-Hz resolution, of EDA, HR, and HRV. For STRT data, values over 2 s were considered outliers and removed. For the recognition test, the percentage of hits for each participant and story was calculated. In the case of the cued and free recall tests, eight participants did not take the phone call, and thus the data contain the responses from 33 participants. Two researchers evaluated each response of the participants to the tests, and rated it as correct or incorrect. Inter-coder agreement was good in both the cued and free recall tests (Krippendorff’s α of 0.82 and 0.98, respectively). In the cases in which the two coders disagreed, only responses assessed as correct by both coders were considered a valid correct response. Therefore, the data considered in the analyses are: for STRT, the response times for each beep for each participant in each story; for the recognition test, the percentage of correct responses of each participant for each story; for the cued recall test, a binary value (0/1) describing whether

Even though a priori statistical power tests were not conducted, the within-subject experimental design applied, in which each participant watched 10 stories per condition (personal/professional), with 41 participants, gives a sample of 820 viewings (410 per condition), which was considered a reasonable sample size.

Journal of Media Psychology (2020), 32(1), 40–45

Ó 2019 Hogrefe Publishing

M. Barreda-Ángeles et al., Cognitive Processing of Personal News

the participant provided or not a correct response to each question about each story; and, for the cued recall test, a binary value for each story for each participant, describing whether or not the participant recalled that story.

ratings for each story (resulting from the average values from the pretest), story pace, time (ranging from the first to the last measure for each of the story), and participant’s gender and age. Model 2 (M2) was similar to M1, but included the type of story as a fixed factor. M1 and M2 were compared using a likelihood test, in order to test the effects of type of story over the dependent variables while controlling for the effects of the rest of possible covariates. For the STRT, recognition, cued recall, and free recall, the same approach was followed, with the exceptions that, instead of time, the temporal position of beeps was included in the STRT models, and no variable reflecting time was included for the recognition, free, and cued recall models.

Data Analysis In this section, the procedure for the analyses of the data for the different dependent variables is described. The intraclass correlation coefficients (ICC) suggested that in all cases a significant proportion of the variance in all the dependent variables was explained by individual differences (ICC-EDA = 0.90; ICC-HR = 0.78; ICC-HRV = 0.97; ICC-STRT = 0.37; ICC-Recognition = 0.29; ICC-Cued recall = 0.57; ICC-Free Recall = 0.56). Mixed multilevel models were used for the analysis, and logistic mixed multilevel models in the case of the cued and free recall data. For each variable, a null model without fixed factors was fitted including a random intercept for participants. In the cases in which they significantly improved model fitting according to a likelihood test, a random intercept for celebrities per participant (since different individuals may have different attitudes toward different celebrities) and an autocorrelation function for modeling error structure were added to the models. This was the case of the models for EDA, HR, HRV, and STRT. Departing from the null model with random terms and autocorrelation of errors, for each psychophysiological variable (EDA, HR, and HRV), two more models were fitted. Model 1 (M1) consisted in adding to the null model the following variables as fixed factors: arousal and valence

Results Including the type of story variable significantly improves model fitting for STRT, w2(1) = 5.26; p = .02, cued recall, w2(1) = 12.32; p < .001, free recall, w2(1) = 12.25; p < .001, HR, w2(1) = 4.19; p = .04, and HRV, w2(1) = 43.1; p < .001, models. The coefficients shown in Table 1 indicate that personal stories are associated with significantly higher STRT, better cued and free recall, lower HR, and higher HRV, which provides support for H1, H3, H4, H6, and H7, respectively. However, including the type of story variable did not improve model fitting in the cases of EDA, w2(1) = 0.92; p = .33, and recognition data, w2(1) = 0.53; p = .46, and the coefficients for this variable are not significant in either EDA or recognition models (Table 1). Therefore, H2 and H5 were not supported.

Table 1. Summary of Model 2 for all the dependent variables STRT

Recognition

Cued recall

Free recall

EDA

HRV

Fixed factors Intercept Type of story

502.05*** 7.61*

0.81***

3.02***

1.1

5.54

75.31***

0.01

0.48***

0.71***

0.03

0.32*

540.11*** 7.21***

0.09***

0.47***

0.04

0.03

0.37**

8.1***

Arousal

10.4***

Valence

15.66***

0.01

0.54***

0.07

0.04*

0.46***

5.48***

Pace

16.21*

0.06

0.56

0.17

0.37***

0.82**

4.3**

0.002***

0.01***

0.02

Time Beep Age Gender

5.13*** 0.45 18.8

0.01

0.00

0.16

0.03

0.47

0.02

0.36

0.42

0.26

0.43

181.19

0.5

0.01

0.09

0.61

0.57

4.3

10.01

Random terms Celebrity/Participant

15.52

Participant

57.21

Residuals n observations n groups (celebrity/participant) n groups (participants)

95.59 5,474

0.22 820

0.46 1,320

0.44 660

410 41

1.16

5.24

0.016 200.79 35.79

55,086

54,622

46,758

389

388

Note. The coding for Type of story is 0 = professional, 1 = personal. Coefficients of the STRT, recognition, free recall, cued recall, EDA, HR, and HRV models are expressed, respectively, in ms, percent of hits, log-odds ratio, log-odds ratio, μS, bpm, and ms. EDA = electrodermal activity. HR = heart rate. HRV = heart rate variability. STRT = secondary task reaction time. *p < .05. **p < .01. ***p < .001.

Ó 2019 Hogrefe Publishing

Journal of Media Psychology (2020), 32(1), 40–45

M. Barreda-Ángeles et al., Cognitive Processing of Personal News

Regarding the control variables, arousal was associated with significant increases in STRT, recognition, cued and free recall, and HRV, and with decreases in HR, while valence produced significant increases in STRT, cued recall, EDA, and HR, and decreases in HRV. Pace was positively associated with EDA, and negatively associated with STRT, HR, and HRV, and did not have significant effects on recognition and recall. Time was associated with significant decreases in EDA and HR, but not in HRV. Age did not show any significant effect on any of the variables, while gender only had a significant effect on HRV.

attributable to individual dispositions (e.g., in terms of fandom or parasocial relationships). The effects of personal stories were significant even when controlling for aspects of the messages such as arousal, valence, and pace, and individual characteristics such as age and gender. One possible limitation is that, according to the pretest ratings, personal stories were more negatively valenced, but the fact that valence was controlled in the models dismisses the possible confounding of these variables in our results. Furthermore, according to the pretest scores, the difference in personal information between personal and professional stories was much larger (Cohen’s d = 5.19) than the difference in valence (Cohen’s d = 1.11). Therefore, it is unlikely that the effects found for the type of story are attributable to a confusion with valence. Taken together, our results support the view that higher levels of personal information have specific effects on attention and memory, regardless of other attributes of the message or the audience. This of course does not preclude that individual differences, as well as other facets of the message, may moderate these effects; in fact, how this general effect is shaped by those possible interactions is a fruitful trail for future research. Some limitations of this study are that we only tested the effects of personal information of celebrities, and that the measurement of personal information in the pretest may oversimplify the possible different dimensions underlying the concept. Further research should extrapolate these results to messages related to anonymous people, examine more deeply the concept of personal information in media, and also examine other aspects not regarded here, such as the possible effects of the coincidence (or lack of) in gender between the depicted character and the viewer. Other limitations are that we did not assess aspects such as whether personal information produces specifically appetitive or aversive activation, or if visual recognition may produce dissimilar results to the ones found on auditory recognition. We did not examine other possibly different effects of personal and professional information over the duration of the viewing; the role of previous knowledge about the celebrities on information processing; as well as that other possible explanations for the fact that we found significant effects of personal information in storage and retrieval, but not in recognition, which might be due, for instance, to a ceiling effect (as suggested by the high value intercept of the recognition model, indicating a mean recognition of above 80% of the auditory excerpts, regardless of the type of story).

Discussion The rationale underlying this study was that, if personal information has motivational relevance, it should produce similar effects on cognitive processing as emotionally arousing contents do. The results support this assumption only partially. We found that personal stories enhance the allocation of cognitive resources to the processing, are better stored and retrieved from memory, and require more resources to be encoded. However, personal news did not produce any significant effect on EDA levels or information recognition. This suggests that, although personal information may not have motivational relevance in the same sense that, for instance, other arousing contents, it has clear effects on attention and memory for the information. The significant effects on HRV and the lack of effects on EDA indicate that only the activity of the parasympathetic nervous system was modulated by personal information. Since parasympathetic activity is associated with attentional focus and information intake, while sympathetic activity is related to mobilization and action preparation, we suggest the following putative explanation for these results: Personal information is important for the social group, and hence deserves attention (leading to parasympathetic activation), but it is unlikely to require an immediate action (like stimuli related to violence or sex, for instance, may require), therefore there is no need for sympathetic activation. On the other hand, the lack of effects of type of story on recognition, together with its significant improvements on cued and free recall, suggest that personal information does not increase the quantity of information encoded, but merely improves how it is stored and retrieved from memory. In order to control for individual differences in disposition toward celebrities, random intercepts for participants and for celebrities per participant were included in all models when they improved model fitting. Moreover, the sample contained stories from 10 different celebrities. Hence, the effects found in the analysis can be considered as general effects in the sample of participants and are not Journal of Media Psychology (2020), 32(1), 40–45

Conclusion Despite its limitations, the present work provides preliminary evidence of the specific effects of personal information Ó 2019 Hogrefe Publishing

M. Barreda-Ángeles et al., Cognitive Processing of Personal News

from a cognitive and experimental perspective, which should be expanded by future work to explore the aforementioned aspects as well as other implications. For example, in the context of political information, it would be interesting to address the question of whether the presence of details on personal facets of political candidates, since it might increase attention, might also increase memory for political information (and not only for personal information).

D. O. Braithwaite (Eds.), Engaging theories in interpersonal communication (pp. 309–322). Thousand Oaks, CA: Sage. Turner, G. (2004). Understanding celebrity. Thousand Oaks, CA: Sage. History Received December 12, 2017 Revision received September 18, 2018 Accepted October 4, 2018 Published online March 8, 2019 Acknowledgments We thank Robert Bailey for language review.

Research Transparency Statement The authors are not willing to share their data, analytics methods, and study materials with other researchers.

References Baumeister, R. F., Zhang, L., & Vohs, K. D. (2004). Gossip as cultural learning. Review of General Psychology, 8(1), 111–121. https://doi.org/10.1037/1089-2680.8.2.111 Bradley, M. M., Codispoti, M., Cuthbert, B. N., & Lang, P. J. (2001). Emotion and motivation I: Defensive and appetitive reactions in picture processing. Emotion, 1(3), 276–298. https://doi.org/ 10.1037/1528-3542.1.3.276 Carpenter, A., & Greene, K. (2016). Social penetration theory. In C. R. Berger & M. E. Roloff (Eds.), The international encyclopedia of interpersonal communication (1st ed., pp. 1–4). Hoboken, NJ: Wiley-Blackwell. Dunbar, R. I. M. (2004). Gossip in evolutionary perspective. Review of General Psychology, 8, 100–110. https://doi.org/10.1037/ 1089-2680.8.2.100 Foster, E. K. (2004). Research on gossip: Taxonomy, methods, and future directions. Review of General Psychology, 8(1), 78–99. https://doi.org/10.1037/1089-2680.8.2.78 Keene, J. R., Clayton, R. B., Berke, C. K., Loof, T., & Bolls, P. D. (2017). On the use of beats-per-minute and interbeat interval in the analysis of cardiac responses to mediated messages. Communication Research Reports, 34(3), 265–274. https://doi. org/10.1080/08824096.2017.1334640 Laborde, S., Mosley, E., & Thayer, J. F. (2017). Heart rate variability and cardiac vagal tone in psychophysiological research: Recommendations for experiment planning, data analysis, and data reporting. Frontiers in Psychology, 8(213). https://doi.org/ 10.3389/fpsyg.2017.00213 Lang, A. (2006). Using the limited capacity model for motivated mediated message processing to design effective cancer communication messages. Journal of Communication, 56(s1), s57–s80. https://doi.org/10.1111/j.1460–2466.2006.00283.x Lang, A., Park, B., Sanders-Jackson, A. N., Wilson, B. D., & Wang, Z. (2007). Cognition and emotion in TV message processing: How valence, arousing content, structural complexity, and information density affect the availability of cognitive resources. Media Psychology, 10(3), 317–338. https://doi.org/10.1080/15213260701532880 Peirce, J. W. (2007). PsychoPy–psychophysics software in Python. Journal of Neuroscience Methods, 162(1–2), 8–13. https://doi. org/10.1016/j.jneumeth.2006.11.017 Petronio, S., & Durham, W. (2008). Understanding and applying communication privacy management theory. In L. A. Baxter &

Ó 2019 Hogrefe Publishing

ORCID Miguel Barreda-Ángeles https://orcid.org/0000-0002-5056-7633 Miguel Barreda-Ángeles Eurecat – Technology Centre of Catalonia C/ Bilbao, 72 08018 Barcelona Spain miguel.barreda@eurecat.org

Miguel Barreda-Ángeles (PhD, 2014) is a researcher at the Digital Humanities Unit at Eurecat – Technology Centre of Catalonia, Spain.

Julia Virgós (MSc, 2017) is currently a clinical psychologist at the Psychiatry and Psychology Unit at Psima and at the Psychooncology Unit at Instituto Médico Tecnológico, Barcelona, Spain.

Maddalena Fedele (PhD, 2011) is a Juan de la Cierva Fellow at the School of Communication and International Relations, Blanquerna – Ramon Llull University, Barcelona, Spain.

Alexandre Pereda-Baños (PhD, 2008) is head of the research line in Perception and Cognition at Eurecat – Technology Centre of Catalonia, Spain.

Journal of Media Psychology (2020), 32(1), 40–45

Call for Papers “Theoretical Consolidation and Innovation in Media Psychology” A Special Issue of the Journal of Media Psychology, Curated by the Editors Nicholas D. Bowman,1 Allison Eden,2 Jesse Fox,3 Christoph Klimmt,4 Özen Odağ,5 Ivar Vermeulen,6 and Catalina Toma7 1

Texas Tech University, Lubbock, TX, USA

Michigan State University, East Lansing, MI, USA

The Ohio State University, Columbus, OH, USA Department of Journalism and Communication Research (IJK), Hanover University of Music, Drama, and Media, Hannover, Germany

4 5

Touro College, Berlin, Germany

University of Wisonsin-Madison, Madison, WI, USA

Vrije Universitteit Amsterdam, The Netherlands

Background and Mission Media psychologists constantly expand the range of domains, themes, and questions that they address in their research. Technological innovations – digitalization, mediatization, and dynamics of convergence in particular – motivate scholars to carry out many empirical studies on users, their processing, their experience, and (co-)production of messages, and related psychological outcomes. It is the mission of peer-reviewed journals such as the Journal of Media Psychology (JMP) to serve as chronicle of this progress and as archive of the knowledge produced. However, the strong research dynamics and persisting arrival of fresh empirical insights can only lead to a flourishing and effective academic field if theoretical work moves ahead as well. Much empirical work in media psychology is guided by existing general theoretical frameworks (e.g., from social psychology or cognitive psychology); other studies rest on field-specific theoretical approaches (e.g., parasocial interaction, narrative persuasion). But regardless the origin of the theoretical base that media psychologists build on, it is important for the field to accompany the empirical day-to-day research business by a reflection on where a given theoretical approach is standing. The mission of the special issue that the Editors of JMP will be curating and that will appear in early 2021 is to Journal of Media Psychology (2020), 32(1), 46–47 https://doi.org/10.1027/1864-1105/a000277

catalyze such theoretical development by mobilizing authors’ efforts for either formalizing new models or theories in media psychology or for discussing prominent existing theoretical frameworks, their empirical validation, predictive usefulness, and potential needs for revision, extension, or even retirement. To serve the building of a community consensus over the key theoretical resources of the field, these discussions should therefore either focus on consolidation (i.e., review of existing works and debates around a theory or model) or innovation (i.e., propositions of extensions to or modifications of an existing theory or model, or even the development of entirely new conceptual approaches). We thus call for manuscripts that focus on a specific theory or model in media psychology. This theory or model is: either one that already exists and is “in use” and for which conceptual reviewing and knowledge consolidation or innovation/modification is indicated, or one that the authors develop as innovative addition to the inventory of theories and models of the field. Such theoretical progress (through consolidation or innovation) will be most likely to materialize if authors team up who have been working on similar themes in the past but have not (yet) joined their theoretical perspectives. It is our explicit intention to inspire the formation of such new authorship networks, as we Ó 2020 Hogrefe Publishing

Call for Papers

(a) want to increase the probability of actually new original thinking, and (b) want to avoid self-promoting review contributions of single researchers or research groups who essentially summarize and promote their own past (theoretical) work. The idea of forming new authorship teams is not mandatory, but strongly recommended in order to achieve effective theoretical consolidation or innovation.

Manuscript Preparation and Submission Only theoretical contributions that comply with the mission statement will be considered for this special issue. Authors are kindly asked to submit their manuscripts as Special Issue Articles through JMP’s editorial management system at http://www.editorialmanager.com/jmp However, deviating from normal principles, submissions to this special issue will be subject to a two-stage process. In the first stage, authors are kindly invited to hand in an extended abstract of the theoretical consolidation or innovation project that they intend to complete. These abstracts should summarize and describe the envisioned content of the future full paper, and enable readers to understand the focus of theoretical consolidation or innovation as well as the line of argumentation and architecture of reasoning. An outlook of how the full text will be organized and structured is also a mandatory element of the extended abstract.

Ó 2020 Hogrefe Publishing

These abstracts must not exceed 2,000 words (excluding references). In order to reduce the risk for authors, selection decisions will be made jointly by the entire team of JMP editors, based on external reviews of these abstracts. Only successful abstract submitters will be requested to hand in full manuscripts in the second stage. For these full submissions, publication in the special issue will be guaranteed, potentially after revision(s) according to editorial feedback and advice. All manuscripts should be prepared in accordance with the journal’s author guidelines for “Theoretical Articles”: http://www.hogrefe.com/periodicals/journal-of-mediapsychology/

Timeline Submission of extended abstracts by April 30, 2020 Invitations to authors of accepted abstracts to submit full papers by June 30, 2020 Deadline for the submission of full papers October 31, 2020 The Editor-in-Chief will oversee the entire process of special issue operations and also respond to any author inquiries. Please direct your correspondence to Christoph Klimmt E-mail christoph.klimmt@ijk.hmtm-hannover.de We wish all authors good luck, fun with finding co-authors, and much inspiration for theorizing and writing!

Journal of Media Psychology (2020), 32(1), 46–47

Instructions to Authors Journal of Media Psychology (JMP) is committed to publishing original, high-quality papers which cover the broad range of media psychological research. This peer-reviewed journal focuses on how human beings select, use, and experience various media as well as how media (use) can affect their cognitions, emotions, and behaviors. It is also open to research from neighboring disciplines as far as this work ties in with psychological concepts of the uses and effects of the media. In particular, it publishes multidisciplinary papers that reﬂect a broader theoretical and methodological spectrum and comparative work, e.g., cross-media, cross-gender, or crosscultural. As JMP is intended to foster Open Science Practices, authors are offered to publish their data and materials (i.e., stimuli and surveys) as Electronic Supplementary Material on the publisher’s website at http://econtent@hogrefe. com. In line with the Peer Reviewers’ Openness Initiative, authors may be asked by reviewers to share their data and materials at any stage of the reviewing process. Journal of Media Psychology publishes the following types of article: Original Articles, Theoretical Articles, Research Reports, Pre-Registered Reports.

Manuscript submission: All manuscripts should in the first instance be submitted electronically at http://www.editorialmanager.com/ jmp. Detailed instructions to authors are provided at https:// www.hogrefe.com/j/jmp Copyright Agreement: By submitting an article, the author confirms and guarantees on behalf of him-/herself and any coauthors that the manuscript has not been submitted or published elsewhere, and that he or she holds all copyright in and titles to the submitted contribution, including any figures, photographs, line drawings, plans, maps, sketches, and tables, and that the article and its contents do not infringe in any way on the rights of third parties. The author indemnifies and holds harmless the publisher from any third party claims. The author agrees, upon acceptance of the article for publication, to transfer to the publisher the exclusive right to reproduce and distribute the article and its contents, both physically and in nonphysical, electronic, or other form, in the journal to which it has been

Journal of Media Psychology (2020), 32(1)

submitted and in other independent publications, with no limitations on the number of copies or on the form or the extent of distribution. These rights are transferred for the duration of copyright as defined by international law. Furthermore, the author transfers to the publisher the following exclusive rights to the article and its contents: 1. The rights to produce advance copies, reprints, or offprints of the article, in full or in part, to undertake or allow translations into other languages, to distribute other forms or modified versions of the article, and to produce and distribute summaries or abstracts. 2. The rights to microfilm and microfiche editions or similar, to the use of the article and its contents in videotext, teletext, and similar systems, to recordings or reproduction using other media, digital or analog, including electronic, magnetic, and optical media, and in multimedia form, as well as for public broadcasting in radio, television, or other forms of broadcast. 3. The rights to store the article and its content in machinereadable or electronic form on all media (such as computer disks, compact disks, magnetic tape), to store the article and its contents in online databases belonging to the publisher or third parties for viewing or downloading by third parties, and to present or reproduce the article or its contents on visual display screens, monitors, and similar devices, either directly or via data transmission. 4. The rights to reproduce and distribute the article and its contents by all other means, including photomechanical and similar processes (such as photocopying or facsimile), and as part of so-called document delivery services. 5. The right to transfer any or all rights mentioned in this agreement, as well as rights retained by the relevant copyright clearing centers, including royalty rights to third parties. Online Rights for Journal Articles: Guidelines on authors’ rights to archive electronic versions of their manuscripts online are given in the ‘‘Guidelines on sharing and use of articles in Hogrefe journals’’ on the journal’s web page at https://www. hogrefe.com/j/jmp October 2016

Ó 2020 Hogrefe Publishing

Psychological Test Adaptation and Development Official Open Access Organ of the European Association of Psychological Assessment (EAPA)

“PTAD will be an important outlet for everyone interested in assessment!” Mathias Ziegler, Editor-in-Chief, Humboldt University Berlin

Volume 1 / Number 1 / 2020

Psychological Test Adaptation and Development

Editor-in-Chief Matthias Ziegler

Official Open Access Organ of the European Association of Psychological Assessment

OA New al Journ

About the journal PTAD is the first open access, peer-reviewed journal publishing papers which present the adaptation of tests to specific needs (e. g., cultural), test translations or the development of existing measures. Moreover, the focus is on the empirical testing of the psychometric quality of these measures. The journal provides a paper template, and registered reports are strongly encouraged. It is a unique outlet for research papers portraying adaptations (e. g., translations) and developments (e. g., state to trait) of individual tests – the backbone of assessment. The expert editor-in-chief is supported by a stellar cast of internationally renowned associate editors. A generous APC waiver program is available for eligible authors. Benefits for authors: • Clear guidance on the structure of papers helps you write good papers • Fast peer-review, aided by the clear structure of your paper • With the optional registered report format you can get expert advice from seasoned reviewers to help improve your research • Open access publication, with a choice of Creative Commons licenses • Widest possible dissemination of your paper – and thus of qualified information about your test and your research • Generous APC waiver program and discounts for members of selected associations The journal welcomes your submissions! All manuscripts should be submitted online via Editorial Manager, where full instructions to authors are also available: https://eu.hogrefe.com/j/ptad

Help medical and other health care students prepare for behavioral science examinations “Targeted specifically for medical students with the goal of helping them pass the behavioral science section of the USMLE, this book is extraordinarily useful for all those interested in the interaction of behavior and medicine.” Bradley R. Cutler, MD, Rush University Medical Center, in Doody’s Book Reviews, (reviewing of the 5th edition)

Danny Wedding, PhD / Margaret L. Stuber, MD (Editors)

Behavior and Medicine 6th edition 2020, xvi / 354 pp. US $69.00 / € 59.95 ISBN 978-0-88937-560-4 The latest edition of this popular textbook on the behavioral and social sciences in medicine has been fully revised and updated to meet the latest requirements on teaching recommended by the National Academy of Medicine (NAM). It is an invaluable resource for behavioral science foundation courses and exam preparation in the fields of medicine and health, including the USMLE Step 1. Its 23 chapters are divided into five core sections: mind– body interactions in health and disease, patient behavior, the physician’s role, physician–patient interactions, and social and cultural issues in health care. Under the careful guidance and editing of Danny Wedding, PhD, Distinguished Consulting Faculty Member, Saybrook University, Oakland, CA, and Margaret L. Stuber, MD, Professor of Psychiatry and Biobehavioral Sciences at UCLA, nearly 40 leading educators from major medical

www.hogrefe.com

Out h Marc 2020

faculties have contributed to produce this well-designed textbook. The following unique features of Behavior and Medicine make it one of the most popular textbooks for teaching behavioral sciences: • Based on the core topics recommended by the NAM • Numerous case examples, tables, charts, and boxes for quick access to information • Resources for students and instructors, including USMLE-style review Q & As • Specific “Tips for the Step” in each chapter guide learning • The use of works of art, poetry, and aphorisms “humanize” the material • C omprehensive, trustworthy, and up-to-date • C ompetitive price