Tesis Evaluacion de la calidad de vida en servicios sociales by laura gomez

UNIVERSITY OF SALAMANCA, FACULTY OF PSYCHOLOGY INSTITUTE ON COMMUNITY INTEGRATION (INICO)

DOCTORAL THESIS-DOCTOR EUROPEUS

“Assessment of Quality of Life in Social Services: Validation and Calibration of the GENCAT Scale”

Author: Laura E. Gómez Sánchez Advisers: Miguel Ángel Verdugo Alonso Benito Arias Martínez

Salamanca, 2010

INDEX INDEX OF CONTENTS CHAPTER 1. INDEX

INDEX OF TABLES

INDEX OF FIGURES

CHAPTER 1.

THE CONCEPT OF INDIVIDUAL QUALITY OF LIFE

CHAPTER 2.

THE ASSESSMENT OF INDIVIDUAL QUALITY OF LIFE

2. 1. Evaluation instruments for people with intellectual disability

2. 2. Evaluation instruments for people with physical disability

2. 3. Evaluation instruments for people with sensory disability

2. 4. Evaluation instruments for people with mental health problems

2. 5. Evaluation instruments for people with substance abuse

2. 6. Evaluation instruments for people with HIV/AIDS

2. 7. Evaluation instruments for elderly people

3. 1. Goal

3. 2. Procedure

3. 2. 1. Building a pool of items

3. 2. 2. Seeking expert opinion

3. 2. 2. 1. Goal...............................................................................................23 3. 2. 2. 2. Participants....................................................................................24 3. 2. 2. 3. Procedure......................................................................................25 3. 2. 2. 4. Results...........................................................................................25 3. 2. 3. Initial version of the GENCAT Scale

3. 2. 4. Consulting Focus Groups

3. 2. 4. 1. Goal...............................................................................................29

3. 2. 4. 2. Participants....................................................................................30 3. 2. 4. 3. Procedure......................................................................................30 3. 2. 4. 4. Results...........................................................................................30 3. 2. 5. Pilot version of the GENCAT Scale

3. 2. 6. Translation of the GENCAT Scale into Catalan

3. 2. 6. 1. Translation into Catalan................................................................31 3. 2. 6. 2. Back-translation............................................................................32 3. 2. 6. 3. Committee of experts....................................................................32 3. 2. 6. 4. Testing the translated version.......................................................32 CHAPTER 4.

VALIDATION BY MEANS OF CLASSICAL TEST THEORY (CTT)

AND CONFIRMATORY FACTOR ANALYSIS (CFA)

4. 1. Goal

4. 2. Method

4. 2. 1. Sampling procedure

4. 2. 2. Participants

4. 2. 3. Instrument

4. 2. 4. Application procedure

4. 3. Results

4. 3. 1. Validity

4. 3. 1. 1. Validity evidence based on scale content.....................................39 4. 3. 1. 2. Validity evidence based on internal structure...............................40 42 The second model tested was the one by Schalock & Verdugo (2002): quality of life comprises eight dimensions that correlate with each other. The factorial design of this model (Figure 2) shows that the prediction errors (θ) range between .20 (p6_2) and .86 (p2_2). Therefore, the coefficients of determination (R2) (Table 1) range between .14 and .80. Half these values are above .50. Regarding factor loadings (λ), they fall within the range of between .38 (p2_2) and .89 (p6_2). 81.3% of the factorial saturations exceed the value of .50 and they were all statistically significant with t values of more than 2.58 (p< .01). Furthermore, all the coefficients were significant.

43 4. 3. 2. Reliability of the scale

4. 4. 2. 1. Internal consistency and standard error of measurement..............55 4. 4. 3. Reliability of the items

CHAPTER 5.

VALIDATION OF THE SCALE USING

MULTI-GROUP EXPLORATORY STRUCTURAL EQUATION MODELING

AND CALIBRATION USING THE RASCH MODEL

5. 1. Goals and hypotheses

5. 2. Method

5. 2. 1. Participants

5. 2. 2. Instrument

5. 2. 3. Procedure

5. 3. Results

5. 3. 1. Evaluation of factor invariance

5. 3. 2. Analysis with the Item Response Theory (IRT) Model

CHAPTER 6.

THE QUALITY OF LIFE OF USERS OF SOCIAL SERVICES IN CATALONIA

6. 1. Overall goal and hypotheses

6. 2. Method

6. 2. 1. Participants

6. 2. 2. Instrument

6. 2. 3. Procedure

6. 3. Results

CHAPTER 7.

DISCUSSION AND FUTURE LINES OF RESEARCH

7. 1. Specific conclusions 7. 1. 1. Building a quality of life scale

78 79

7. 1 .1. 1. Building a pool of items ...............................................................79 7. 1. 1. 2. Seeking expert opinion..................................................................80 7. 1. 1. 3. Discussion groups.........................................................................81

7. 1. 1. 4. Translation of the scale into Catalan.............................................81 7. 1. 2. Validation of the GENCAT Scale by CTT and CFA

7. 1. 2. 1. Sample selection...........................................................................82 7. 1. 2. 2. Participants and procedure............................................................83 7. 1. 2. 3. Validity evidence based on content..............................................83 7. 1. 2. 4. Validity evidence based on internal structure...............................83 7. 1. 2. 5. Reliability of the scale and subscales............................................84 7. 1. 2. 6. Reliability of items........................................................................85 7. 1. 3. Validation of the Scale using multigroup ESEM

7. 1. 4. Calibration of the GENCAT Scale with the RSM

7. 1. 5. The quality of life of the users of social services in Catalonia

7. 2. General conclusions

7. 2. 1. The study’s strong points

7. 2. 2. The study’s limitations

7. 2. 3. Future lines of research

REFERENCES

APENDIX 1.

111

ENGLISH VERSION OF THE GENCAT SCALE

111

INDEX OF TABLES Table 1. Factor loadings (λ), prediction errors (θ) and coefficients of determination (R2) for model 2

Table 2. Fit indices for the various models

Table 3. Composite reliability (ρc) and average variance extracted (ρv)

Table 4. Coefficients of correlation between the model’s latent variables

Table 5. Fit indices for the eight-dimension model in elderly people and in those at social disadvantage (ESEM)

Table 6. Fit indices for the multigroup invariance analysis without restrictions

Table 7. Descriptive statistics of the scores on the GENCAT Scale

INDEX OF FIGURES Figure 1. Standardised parameters for model 1 (unidimensional)

Figure 2. Standardised parameters for model 2 (Schalock & Verdugo, 2002)

Figure 3. Standardised parameters for model 3 (Wang et al., in press)

Figure 4. Standardised parameters for model 4 (Salamanca model)

Figure 5. Standardised parameters for model 5 (Schalock model)

Figure 6. Model for elderly people and for those at social disadvantage

Figure 7. Regions of Catalonia

CHAPTER 1. THE CONCEPT OF INDIVIDUAL QUALITY OF LIFE

The overriding purpose of this chapter is to answer the question of what we understand individual quality of life to be. Accordingly, we begin by presenting some brief historical notions on the concept’s development. These reveal that prior to the 1960s only approximate concepts were to be found that contributed to the birth of the quality of life construct, but which focused on the regulatory and objective evaluations of concepts that are now considered to be quite different. It was not until the end of the 1960s and during the 70s that the concept was effectively linked to a notion that incorporated subjective evaluations that included aspects such as personal feelings of happiness or satisfaction, and only in the 80s did the concept become consolidated with operationalisation and evaluation beginning with different models. Nonetheless, the 90s were the decade of results and the disclosure of findings. The concept was forged when much more detailed work was carried out on the specification and breakdown of the concept and its dimensions, its measurement and its integration within professional practices. Research began to increase rapidly during these years, giving rise to more than 100 definitions (Cummins, 1997) and over 1,000 evaluation instruments (Hughes & Hwang, 1996). Growing interest in the subject meant that this decade saw the statement of the so-called 12 principles of quality of life, which seek to respond to the pressing need to clarify the concept and which were proposed and disseminated by an international team of researchers and quality of life professionals (Schalock et al., 2002). Based on their work, the decade we are now in has witnessed the emergence of a growing consensus regarding four approaches to the use of quality of liferelated personal outcomes (Schalock, Gardner & Bradley, 2007).

This overview of the concept’s historical development leads us to today’s understanding of quality of life, in which there is a prevalence of operational models and in which, thanks to the work undertaken by the Special Interest Research Group on Quality of Life of the International Association for the Scientific Study of Intellectual Disabilities (IASSID), we can affirm that there is international consensus on its essential aspects (Schalock & Verdugo, 2008). Amongst the models of individual quality of life that are most widely used, special note should be taken of the Comprehensive Quality of Life Scale by Cummins (1996, 2000, 2005), the model by Felce and Perry (1995, 1996) –also applied to, and adapted for, people with severe and multiple disability (Petry, Maes & Vlaskamp, 2005, 2007), and the model proposed by Schalock and Verdugo (2002, 2007, 2008) –the one most cited today and which seems to be the one most generally accepted by the scientific community both in Spain and abroad. Let us now delve further into the conceptualisation of quality of life according to the multidimensional model by Schalock and Verdugo (Schalock & Verdugo, 2002; 2007; 2008; Schalock, Keith, Verdugo & Gómez, in press; Verdugo, 2006), into its measurement principles and its main uses (Brown, Schalock y Brown, 2009; Schalock & Verdugo, 2007; Tamarit, 2005). In addition, we shall describe the model’s development process by alluding to two main stages that include: (a) its formulation (Aznar & Castañón, 2005; Chou & Schalock, 2009; Cummins, 1997; Felce & Perry, 1997; Gardner & Carran, 2005; Hughes et al., 1995; Hughes & Hwang, 1996; Jenaro et al., 2005; Keith, 2007; Keith, Heal & Schalock, 1996; Schalock & Bonham, 2003; Schalock, Gardner & Bradley, 2007; Schalock & Keith, 1993; Schalock & Verdugo, 2002; Schalock et al., 2005; van Loon, van Hove, Schalock & Claes, 2008; Verdugo, Arias, Gómez & Schalock, 2009; Xu, Wang, Xiang & Hu, 2005); (b) validation of the conceptual and measurement framework through the verification of its factor structure (Bonham, Basehart & Marchand, 2003; Bonham et al., 2004; Jenaro et al., 2005; Martín, 2006; Verdugo, Arias, Gómez y Schalock, in press; Wang, Schalock & Verdugo, in press) and determining the etic and emic properties of its dimensions and indicators (Chou & Schalock, 2009; Jenaro et al., 2005; Keith & Schalock, 2000; Kober & Eggleton, 2002; Rapley & Lobley, 1995; Schalock & Verdugo, 2002; Skevington, 2002; Verdugo, Prieto, Caballo & Peláez, 2005); and (c) implementation of the model through dimensions, indicators and mediating and moderating variables (Schalock et al., in press).

Once the model has been described in detail, we now go on to present this model’s main applications within the sphere of systems and grouped into four main objectives: (a) in the microsystem, the stress is on the evaluation of personal outcomes (e.g., Bonham et al., 2004; Keith, 2007; Schalock, Bonham y Verdugo, 2008; Schalock, Verdugo et al., 2008; van Loon et al., 2008; Verdugo et al., 2007, 2009); (b) in the mesosystem, it is on the creation of the profiles of providers and the development of reports by organisations (Keith & Bonham, 2005; Keith & Ferdinand, 2000) and the design of strategies for improving quality (Bonham et al., 2006; Schalock, Bonham et al., 2008); and (c) in the macrosystem, it is on its use for marking out a reference framework for policies on target results (Shogren et al., in press) and individual support programmes (van Loon, 2008, 2009). The chapter ends with a detail of new challenges and emerging lines in quality of life (Schalock, Gardner & Bradley, 2007): (a) the concept of quality of life as an agent of change; (b) evaluation and feedback regarding quality of life as an integral part of the operations of systems and organisations; (c) users as key players; (d) a redefinition of the role of organisations; (e) new management strategies; and (f) improving quality as a continuous process. Following a review of the literature, we should like to add another challenge or opportunity: (g) the development of a theory of quality of life.

CHAPTER 2. THE ASSESSMENT OF INDIVIDUAL QUALITY OF LIFE

As in the previous chapter, the general aim underpinning this chapter involves responding to the following question: how is individual quality of life evaluated? To do so, we have firstly explained the importance the concept has today (Schalock, Gardner & Bradley, 2007; Schalock & Verdugo, 2002), above all in social, education and health services, as it can be used for several purposes that include: (a) the objective evaluation of people's needs and their subjective levels of satisfaction; (b) the evaluation of the results of those programmes, strategies and activities designed to improve quality undertaken by human and social services; (c) the gathering of relevant information that provides a target and a guide for the provision of services; and (d) the planning and formulation of policies rolled out to enhance the quality of life of people with and without disabilities and the quality of those organisations that provide social services to different groups in a situation of risk or exclusion and which require support. There now follows a detailed study of the four measurement principles for quality of life developed and verified by an international team of researchers, as referred to in the first chapter (Schalock et al., 2002; Schalock et al., 2005; Verdugo et al., 2005). We thus note that these principles indicate that the instruments and efforts used for evaluation can differ from each other depending on the evaluation approach adopted. In order to summarise the numerous studies made, they are often all organised according to the perspective described by Borthwick-Duffy in 1992. According to this perspective, there are three possible theoretical frameworks catering for the majority of the studies on quality of life (Arostegui, 2002; Felce & Perry, 1995; Gómez, 2005; Gómez-Vela & Sabeh, 2000): (a) the first theoretical

framework defines quality of life as the quality of one’s conditions of life; quality of life is defined in terms of the conditions of life experienced by an individual and evaluated objectively by an external evaluator; (b) the second one defines quality of life as personal satisfaction with one’s conditions of life (Veenhoven, 1996), their evaluation is only meaningful if it is the person themself who reports on it (subjective evaluation); and (c) the third theoretical framework is a combination of the two preceding ones; quality of life is therefore defined as the combination of the quality of conditions of life and personal satisfaction, as a merger of objective and subjective aspects. It is this last theoretical framework that appears to provide the setting for most of the studies conducted today and the one applied by the majority of scholars. Within this perspective there is consensus regarding the fact that, depending on the purpose and perspective of the instrument developed, the indicators on quality of life can be used to evaluate the wellbeing perceived by an individual or the experiences and circumstances in the life of that person (Schalock, Bonham et al., 2008). Along these lines, we can refer to evaluations or evaluation instruments that are objective or subjective depending on their purpose, content and respondent (Bonham et al., 2004, 2006; Brown et al., 2004; Cummins, 1997, 2005; Gardner & Carran, 2005; Keith & Bonham, 2005; Keith & Schalock, 2000; Perry & Felce, 1995, 2005; Schalock, Bonham et al., 2008; Schalock & Felce, 2004; Schalock, Gardner & Bradley, 2007; Schalock & Verdugo, 2002, 2007; Schalock et al., 2002; van Loon et al., 2008; Verdugo, Arias & Gómez, 2006; Verdugo et al., in press; Verdugo, Arias, Gómez & van Loon, 2007; Verdugo, Gómez & Arias, 2007, in press; Verdugo, Gómez, Schalock & Arias, in press; Verdugo, Schalock, Gómez & Arias, 2007; Walsh et al., 2006). We shall now conduct a critical review of existing evaluation instruments for each one of the collectives involved in this work. As one might expect, this will show that the sphere of intellectual disability is the one with the largest array of instruments from the perspective of individual quality of life (not surprisingly, it is within this sphere that the aforementioned model of quality of life has emerged and undergone its greatest development).

2. 1. Evaluation instruments for people with intellectual disability Amongst the questionnaires most widely used for the objective evaluation of quality of life are the Program Analysis of Service Systems (PASS) and the Program Analysis of Service Systems’ Implementation of Normalization Goals (PASSING) (Wolfensberger & Glen, 1975; Wolfensberger & Thomas, 1983). Another questionnaire applied extensively was the Living in a Supervised Home: A Questionnaire on Quality of Life (Cragg & Harrison, 1985). A more recent instrument developed in step with the current line of objective and multidimensional evaluations of the individual quality of life of people with intellectual disability is the Evaluation of Quality of Life Instrument (EQLI) (Nota, Soresi & Perry, 2006). Amongst those instruments for measuring the quality of life of people with intellectual disability from a subjective perspective, the ones most widely used are the Lifestyle Satisfaction Scale (LSS) (Heal & Chadsey-Rusch, 1985) and the Multifaceted Lifestyle Satisfaction Scale (MLSS) (Harner & Heal, 1993). We consider both these approaches to be evaluation perspectives that are to be superseded, as an evaluation based exclusively on either subjective or objective aspects provides a limited view and understanding of quality of life. Accordingly, in line with the principles on the measurement of quality of life (Schalock et al., 2002; Verdugo et al., 2005), we propose more holistic and more rounded evaluation approaches that consider both objective and subjective aspects. Nonetheless, as we noted in the previous section, objective or subjective instruments can be used depending on the purpose of the evaluation. Therefore, if the aim is to evaluate a person’s satisfaction, then this may be the appropriate approach. The more holistic, rounded and integrating perspective is the one that combines the evaluation of objective and subjective aspects. Amongst the oldest of these are the Quality of Life Interview Schedule (QUOLIS) (Ouellette-Kuntz, McCreary, Minnes & Stanton, 1994), the Quality of Life Questionnaire (QOLQ) (Schalock & Keith, 1993) and the Comprehensive Quality of Life Scale-Intellectual Disability (ComQol-ID) (Cummins, 1993). These have all been superseded by others of more recent design, such as the Escala Integral de Calidad de Vida (‘Integral Quality of Life Scale’) (Verdugo, Gómez, Arias & Schalock, 2009) and the Personal Outcomes Scale (POS) (van Loon, van Hove, Schalock & Claes, 2008).

There are several authors who coincide in pointing to the scarce or even nonexistent relationship between objective and subjective evaluations (Cummins, 2000, 2005; Gómez, 2005; Janssen, Schuengel & Stolk, 2004; Keith & Schalock, 2000; Perry & Felce, 2005; Verdugo, Arias & Gómez, 2006; Verdugo, Gómez & Arias, 2009; Verdugo, Gómez, Arias & Schalock, 2009). These discrepancies between them impel us to continue our research in this field, and they currently inform the general consensus regarding the pressing need to investigate the best ways of evaluating the concept of quality of life, paying special attention to both objective and subjective circumstances (Anderson & Burckhardt, 1999; Cummins, 1996; Goodley, Armstrong, Sutherland & Laurie, 2003; Schalock & Felce, 2004; Schalock & Verdugo, 2002).

2. 2. Evaluation instruments for people with physical disability The review made of the literature on the quality of life of people with physical disability reveals a major prevalence of approaches based on health-related quality of life (HRQL). Only 3 articles refer to a broader understanding and single out the importance of evaluating other factors over and above those related directly to health (Koch, Rumrill, Roessler & Fitzgerald, 2001; Levasseur, Desrosiers & Tribble, 2008; Renwick, Nourhaghighi, Manns & Rudman, 2003). More specifically, one of these three describes the building of a new instrument: the Quality of Life Profile for Adults with Physical Disabilities (QOLP-PD) (Rudman, Renwick, Raphael & Brown, 1995), which is valid for all kinds of disabilities and is based on a thoroughly developed conceptual framework (Renwick & Brown, 1996) derived from the models designed for adults with intellectual disability and consistent with current principles on the measurement of individual quality of life. Only two attempts have been forthcoming in Spain for evaluating the quality of life of people with physical disability. The first involves an extensive line of research initiated by Aguado et al. which specifically seeks to evaluate coping in people with spinal damage (García, 2006; Rueda, 2001; Rueda, Aguado & Alcedo, 2008) and which uses, amongst others, the Escala Multidimensional de Evaluación de Lesionados Medulares (EMELM)

(‘Multidimensional Scale for Evaluating People with Medullar Injuries’) (Aguado et al., 1994, 1997, 2003). The second line of research is the one pursued by the University of Salamanca’s Institute on Community Integration (INICO), which follows a line of work similar to our own approaches through the application of the QOLQ by Schalock & Keith (1993) to a sample of 209 Mexican participants (Caballo et al., 2005) and in which the instrument’s suitability for use with this collective is readily apparent. Nevertheless, we note that there is no instrument with appropriate or proven psychometric properties for evaluating the individual quality of life in people with physical disability from the perspective we posit.

2. 3. Evaluation instruments for people with sensory disability As with the previous collective, the review of the scientific literature on the quality of life of people with sensory disability has been addressed almost exclusively from the perspective of health-related quality of life. In this case, of the 103 results found in the ERIC, OVID and Health and Psychological Instruments databases, only one instance of research fell outside the sphere of the HRQL and adopts an approach to individual quality of life similar to that considered in this work. Apart from general instruments, special note in that perspective should be taken of several specific instruments, such as the Quality of Life and Vision Function Questionnaire (QOL-VFQ) (Carta et al., 1998; Gothwal, Wright, Lamourex & Pesudovs, 2009), the National Eye Institute 25-Item Visual Function Questionnaire (NEI-VFQ-25) (La Grow, 2007; Mangione et al., 2001), the Vision-Related Quality of Life Core Measure (VCM1) (Frost et al., 1998) and the Low Vision Quality of Life (LVQOL) (Wolffsohn & Cochrane, 2000). From the perspective of individual quality of life, the sole work undertaken is a study of the psychometric properties of the QOLQ by Schalock & Keith (1993) involving a sample of 364 people with visual disability in the Spanish region of Castilla y León (Verdugo et al., 2005). The findings of confirmatory factor analysis showed that this collective's resulting factor structure was not consistent with that of the original scale (Kober & Eggleton, 2002; Rapley & Lobley, 1995).

2. 4. Evaluation instruments for people with mental health problems There are two types of approaches to the evaluation of quality of life in people with mental health issues. The first of these involves the use of generic instruments regarding state of health and the second one involves specific instruments that focus on clinical aspects of particular illnesses. Both approaches in this collective share the dominant approach of health-related quality of life. In general, there is also agreement in this field that quality of life should reflect a person’s wellbeing in both objective and subjective terms (Hewitt, 2007; Navarro, García-Heras, Carrasco & Casas, 2008). Along these lines, note should be taken of two instruments for people with schizophrenia that are fairly close to our own perspective on quality of life. The first of these is the Modular System for Quality of Life (MSQoL) (Pukrop, Möller & Steinmeyer, 2000) and the second one is the Seville Questionnaire of Quality of Life (Giner, Ibáñez, Franco & Alarcón, 2008). Both instruments lack important dimensions in the evaluation of the quality of life of this collective, such as for example self-determination (Verdugo & Martín, 2002). Something similar occurs in the field of depression, we again note the prevalence of healthrelated quality of life models and the evaluation of quality of life using generic and specific instruments developed from that perspective (e.g. Abdel-Kader, Unruh & Weisbord, 2009; Angermeyer, Holzinger, Matschinger & Stengler-Wenzke, 2002; Herrman & Chopra, 2008; Moore, Höfer, McGee & Ring, 2005; Wisniewski et al., 2007).

2. 5. Evaluation instruments for people with substance abuse In the search for instruments for the evaluation of people with some form of substance abuse, and within the field of the evaluation of health-related quality of life, we find a marked tendency to use generic instruments (EuroQol Group, 1990; Hunt et al., 1985; Stewart & Ware, 1992; Ware & Sherbourne, 1992), and only four specific ones for people with substance abuse: Qol-DA, IDUQoL, TECVASP and the HRQOLDA Test. Nonetheless, insofar as the sphere of so-called individual quality of life is concerned, hardly any cases are to be found. In fact, Fischer, Rehm & Kim (2001a, 2001b) already alerted to the pressing need to research the conceptualisation of quality of life based on the perspectives and experiences of people with substance abuse (Farquhar, 1995).

Along these lines, given the almost exclusive attention paid to HRQL in the research into substance abuse, De Maeyer et al. (2009) conducted a study designed to achieve a better understanding of the construct in people with substance abuse through the holding of nine discussion groups in Belgium. The study’s conclusions indicated that people with substance abuse did not predominantly relate their quality of life to health, but instead referred to many other aspects that went beyond those typically included in the evaluations of health-related quality of life. Specifically, they observed that those indicators the participants provided spontaneously could be grouped into the eight dimensions proposed by Schalock & Verdugo (2002) and emphasised the need for an instrument that would allow evaluating individual quality of life from this perspective.

2. 6. Evaluation instruments for people with HIV/AIDS As in the case of the evaluation of the previous collectives, the evaluation of people with HIV and AIDS has so far been addressed exclusively from the perspective of HRQL (e.g., Cunningham, Crystal, Bozzette & Hays, 2005; Heckman, 2003; Ryu et al., 2009). Nevertheless, certain studies made from that perspective coincide in indicating the importance of evaluating other aspects not directly related to health, such as support, satisfaction, emotional wellbeing and interpersonal relations (Heckman, 2003).

2. 7. Evaluation instruments for elderly people Spain is one of the countries in Europe that has recorded the highest rate of ageing among its population. Nevertheless, an ageing population has posed a new challenge and had a major impact on social policies, on professional interests and practices, and on the design and provision of services and support, where the important thing is not only a long life but also a quality one. We might contend that, after the sphere of disability, this is the field in which the concept of individual quality of life is currently acquiring the greatest importance. Two instruments have been found for evaluating the individual quality of life in this field in Spain: (a) the Escala de Calidad de Vida (ECV) (‘Quality of Life Scale’) (Alcedo et al., 2008); and (b) the Escala de Calidad de Vida FUMAT (‘FUMAT Quality of Life Scale’) (Verdugo, Gómez & Arias, in press). However, in short, a review of the scientific literature

clearly shows that there is no quality of life instrument that allows evaluating the quality of life of the users of social services, as we contend in this work.

CHAPTER 3. DEVELOPMENT OF THE GENCAT SCALE

3. 1. Goal The general aim was to develop a quality of life scale for users of social services in Catalonia. In order to guarantee its usefulness, content validity and cultural appropriateness, the GENCAT Scale had to fulfil certain requirements: 1. The scale had to be objective by nature (i.e., evaluate personal outcomes objectively). Thus, the scale should be sensitive to organisational changes, improvement strategies and the services’ care and intervention programmes. It should therefore contain observable aspects of quality of life and be completed by the professionals in those selfsame services. The aim of this requirement is to ensure the instrument's usefulness. 2. The scale should be based on the latest international developments in research into indicators and tests for evaluating quality of life. The purpose behind this requirement was to overcome the limitations found in other instruments, integrate the international state-of-the-art in research into this field and guarantee the innovative nature of the scale. 3. The test should be developed through consensus on indicators amongst the main stakeholders involved in Catalonia. It should therefore be built by considering the

opinions of users of the services, their family members, the professionals attending to them, the representatives of the organisations providing the services and experts from the administration. The aim has thus been to harness the concept’s emic features; that is, those aspects of quality of life that may be specific to Catalan culture. 4. The indicators included in the instrument should be relevant to all the collectives of users of social services (the elderly, people with mental health problems, people with intellectual, physical or sensorial disability, people with substance abuse and people with HIV/AIDS). This has meant reflecting the core principle of quality of life whereby it involves the same dimensions and indicators for everyone, with the focus being on the concept’s universal or etic features. 5. The final requirement involves psychometric properties. The GENCAT Scale should be built with strict methodological rigour and combine quantitative and qualitative methods. The aim has therefore been to imbue the instrument with suitable assurances of reliability and validity.

3. 2. Procedure The development of the GENCAT Scale has followed the recommendations that the American Educational Research Association (AERA), the American Psychological Association (APA) and the National Council on Measurement in Education (NCME) specify as standards for psychological tests, as well as the standards formulated by Prieto & Muñiz (2000) and Carretero-Dios & Pérez (2007).

3. 2. 1. Building a pool of items The process of building the instrument began with a thorough review of the scientific literature on quality of life. Based on the conceptual framework of quality of life proposed by Schalock & Verdugo (2002, 2007, 2008), the review was used to select different indicators and items for each one of the dimensions of quality of life. The majority stemmed

from the developments made of that conceptual framework (Cummins, 1993; Gómez, 2005; Gómez-Vela, 2003; Jenaro et al., 2005; Martín, 2006; Sabeh, 2004; Schalock & Keith, 1993; Schalock & Verdugo, 2007; Schalock et al., 2005; Verdugo, Arias & Gómez, 2006; Verdugo, Gómez, et al., 2009; Verdugo, Schalock, et al., 2005, 2007; Wehmeyer & Kelchner, 1995). This review process provided a pool of 39 indicators and over 200 items. A further screening was made of this pool to discard those items that were redundant or ambiguous and select those of greater relevance. The 39 indicators were then organised into dimensions, selecting at least two relevant items for each one of the indicators and at least 10 items for each one of the eight dimensions, giving a total of 120. All the items were redrafted into the third person singular (he/she) so they could be completed by an external observer. The next step was to organise the items by levels of systems; in other words, according to whether they referred to the microsystem, mesosystem or macrosystem. The 14 items referring to the macrosystem were removed. The reason for their removal was that these items did not reflect personal outcomes that could be targeted by the intervention of a service or organisation, but instead alluded to results directly related to social policies. This meant forming an initial bank of 39 indicators and 106 relevant items for social services, with 68 of them referred to the microsystem and 38 to the mesosystem.

3. 2. 2. Seeking expert opinion 3. 2. 2. 1. Goal

Once it had been verified that the dimensions and indicators of quality of life were represented and that the items allowed evaluating quality of life at the levels of microsystem and mesosystem, the next step involved assessing their appropriateness and representativeness. Guaranteeing the appropriateness and representativeness of the selected indicators and items has likewise validated the scale’s content. This was achieved by seeking expert opinion. The specific aims of this consultation were:

To verify whether the selected items were appropriate or suitable for evaluating the corresponding dimension of quality of life in each collective.

To determine the degree of importance or relevance of the selected items for evaluating quality of life in each collective

To test the items’ degree of observability; that is, the extent to which the actions described in the items were empirically verifiable and could be rated objectively by an external observer.

To check that the main indicators of quality of life were represented or, on the other hand, whether there were other relevant indicators for evaluating the respective dimensions that had not been included.

3. 2. 2. 2. Participants As mentioned earlier, the GENCAT Scale is used to evaluate the quality of life of users of social services. Seven groups or collectives are involved: the elderly, people with mental health problems, people with intellectual, physical or sensorial disability, people with substance abuse and people with HIV/AIDS. For this reason, the participation was sought of experts, or referees, in quality of life and in each one of the seven collectives involved, giving rise to a high number of participants (N = 73). Of these 73 referees, the majority were experts in intellectual disability (n = 18), physical disability (n = 14) and the elderly (n = 12). The group of experts in sensory disability (n = 9) consisted of 3 experts in visual disability and 6 in auditory disability. The group of experts in mental health and substance abuse comprised 8 participants for each of these. The smallest group was the one formed by experts in people with HIV/AIDS (n = 4). Women accounted for 53.42% (n = 39), and men for 46.58% (n = 34). The aim was to involve experts from different Spanish provinces, although a major component would be from Catalonia. 19 Spanish provinces were represented from 11 autonomous communities or regions. The most numerous group was that involving experts from Castilla y León (n = 23), followed by experts from Catalonia (n = 16) and Madrid (n = 15).

3. 2. 2. 3. Procedure The referees were sent an e-mail inviting them to take part, although in many cases a telephone conversation was required to agree upon dates and answer queries. Together with the invitation to take part, the experts were sent a form to fill in certain personal details, a document with the instructions for carrying out the task and a Microsoft Excel ® file with the items to be assessed. The authors of tests tend to base themselves on a specification of the content of the domain, describing it in detail with a clarification of its areas and of the types of items. Accordingly, the items were submitted to the referees organised by dimensions, but without specifying the indicator they were assessing. The instructions provided them with the conceptual framework of quality of life they required for rating the items. Their task involved rating each one of the items on a scale of ‘1’ to ‘10’ for the three criteria mentioned above: appropriateness, importance and observability. These meant awarding three scores to each one of the items, with ‘1’ indicating that the item was very inappropriate, unimportant or unobservable, and ‘10’ indicating that it was very appropriate, important or observable. Furthermore, they were asked to include up to three items or indicators for each one of the dimensions if they deemed them to be relevant and which had not been included in the list. Finally, they were invited to add any comments, suggestions or redrafts of the items they considered convenient.

3. 2. 2. 4. Results The ratings provided by the referees were subjected to quantitative and qualitative concordance analysis. The quantitative analyses were made for the criteria of appropriateness, importance and observability for each one of the collectives involved and for each one of the dimensions of quality of life. The qualitative methodology was used to analyse the additional information provided by the referees (new items and indicators, comments, suggestions and redrafts).

a) Coefficients of concordance between referees In order to analyse the agreement between referees, the original concordance coefficients were calculated (BN) as well as the weighted concordance coefficients (BWN) (Bangdiwala, 1987). These coefficients range between ‘0’ (no agreement) and ‘1’ (full agreement). In most research, weighted coefficients of more than .400 are usually considered suitable. In the case of the original agreements (BN), the coefficients ranged between .090 and .600. Only one coefficient was found lower than .100 in appropriateness and importance. Regarding observability, however, all the coefficients exceeded the threshold of .100. As for the weighted agreements (BWN), the values ranged between .500 and 1.000 for appropriateness and importance and between .400 and 1.000 for observability. The coefficients exceeding .800 accounted for 46.03% and 39.68% for appropriateness and importance, respectively, as opposed to 20.63% recorded by observability. Based on the results forthcoming, we can conclude that, in general terms and with the particularities mentioned, there was a high level of concordance between all the referees in terms of their rating of the items for all the dimensions as highly appropriate (Mean BWN = .761), important (Mean BWN = .769) and observable (Mean BWN = .738).

b) Comparison of the mean of the concordance coefficients in appropriateness, importance and observability by groups of referees Based on the weighted concordance coefficients (BWN) obtained for each one of the groups in the eight dimensions, we calculated the mean agreements for each one of the groups of experts in the different collectives in appropriateness, importance and observability. We used these means to calculate the overall mean agreement for each one of the groups. Although the differences are not statistically significant (F(6) = 1.615; p = .215), the highest concordance coefficients were found amongst the experts in sensory disability (M = .817) and intellectual disability (M = .800). Slightly lower agreement was recorded amongst the experts in the elderly (M = .676) and substance abuse (M = .685). All the other groups returned a mean agreement above .700. Special note should be taken of the peculiarity of the group of experts in HIV/AIDS with a very high mean agreement ( M = . 877) in observability as opposed to an M ≈ .700 in appropriateness and importance.

c) Correlations between appropriateness, importance and observability The referees tended largely to equate the appropriateness of the items with their importance. The correlations obtained were in all cases higher than .800 (p < .001). By contrast, these variables recorded a weaker relationship with observability, providing correlations of around .600 or lower (p < .001). The highest correlations between appropriateness and importance were to be found in the dimensions Self-determination ( ρ = .900), Material wellbeing (ρ = .890), Personal development and Rights (ρ = .890), whereas the lowest between these categories and observability were found in Rights and Social inclusion. The lowest was found in the dimension Interpersonal relations ( ρ = .820). Regarding the lowest correlations found between appropriateness and importance and the observability variable, Rights was the one recording the lowest correlations ( ρ = .560 and ρ = .530, respectively). By contrast, observability correlated highly ( ρ = .850) in the case of appropriateness and moderately so ( ρ = .690) with importance. Self-determination was the dimension with the highest correlations (ρ > .800) between the three variables, followed by Personal development (ρ > .700). d) Comparison of the mean ranges of the items in the dimensions by groups of referees The next step involved calculating the mean ranges of the items in each one of the dimensions, which were compared according to the referees’ specialism by means of the Kruskal-Wallis nonparametric analysis test. This test is an alternative to the one-way analysis of variance F-test for simple classification designs. In this case, a comparison is made between several groups using the median of each one instead of the mean. The referees disagreed on the appropriateness of 13 items: there were significant differences (p < .05) regarding the importance of 14 items and in the observability of 4 items. In short, the results of these concordance analyses involving the experts allow concluding that there were significant discrepancies in the assessment of 22 items. Accordingly, when we consider that the referees assessed 106 items, we can conclude that there was agreement on the appropriateness, importance and observability of 84 items, which means consensus on 79.25% of the total items.

e) Mean ratings and deletion of items In order to safeguard the validity of the content, when selecting the items, priority was given first to the appropriateness that according to the experts each item had for measuring the dimension, followed by its importance for evaluating a person’s quality of life and, finally, its observability. In addition, the decision was taken to discard those items with a widely varying level of agreement. Consequently, a choice was made of those items that fulfilled the following requirements: 1. High scores in appropriateness and importance (M ≥ 8) and little variability (DT < 2.5). 2. Middling scores in observability (M > 6) and little variability (DT < 3). Twenty items failed to meet these criteria (18.86% of the total), so they were discarded. To conclude, the outcome of the referees’ analysis was 86 valid items (81.13% of the total).

f) Selection of the most suitable items A selection was made of the 34 indicators and 55 items considered to be most suitable by the referees and the ones most sensitive to change through organisational strategies or care and intervention programmes.

g) Inclusion of new items and indicators proposed by the referees Regarding the items and indicators proposed by the referees, the first step involved their quantification. In total, 97 new items and 17 new indicators were submitted. The dimension receiving the highest number of contributions was Material wellbeing with 15 items.

A thorough review was made of all the items proposed and nine items were selected that were considered to be the most suitable. The criteria considered for this selection were as follows: (a) the number of referees suggesting the item; (b) the theoretical reasoning for the item’s inclusion in the assigned dimension; (c) the certainty that there were no similar items already included in the initial version of the scale; (d) the applicability to all the collectives involved in the evaluation; (e) the influence on a person’s quality of life; and (f) the possibility of intervention by organisations in the activity involved in the item for improving a person’s quality of life.

3. 2. 3. Initial version of the GENCAT Scale To summarise, 55 items and 34 indicators were chosen from the initial pool. These were accompanied by 9 items and 4 indicators proposed by the referees. This meant that the initial version of the scale consisted of 64 items and 38 indicators.

3. 2. 4. Consulting Focus Groups 3. 2. 4. 1. Goal The next step required the validation of the items and indicators by the people involved in the social services provided by ICASS in Catalonia (etic properties) and by all the groups and collectives targeted by the scale (emic properties). This general aim was embodied in the following specific goals: 1.

Verify that the items and indicators were valid for the users of social services in Catalonia, their family members and the professionals attending to them (emic properties).

Verify that the items and indicators were appropriate, important and observable regarding all the collectives involved in the evaluation (etic properties).

Decide whether there were relevant items and indicators that had not been considered in the initial scale.

Verify whether there were inappropriately drafted or vague items. 29

3. 2. 4. 2. Participants Five focus group sessions were held. They were all chaired by a member of INICO. Each group was made up of several users, their family members and social services staff. Although the aim was for each group to have between 9 and 12 members, they finally consisted of between 7 and 14 people. A total of 54 people took part: 10 users, 11 family members and 33 professionals.

3. 2. 4. 3. Procedure The discussion centred on each dimension one by one. Within each one of the dimensions, in turn, each item and indicator was also discussed. Accordingly, each item and indicator had the following prompting questions: 1.

Do you think the indicator/item affects this collective’s quality of life?

Do you think it is important to measure this indicator/item for this collective?

Do you think the items that measure this indicator can be assessed by an external observer?

Do you think there is an important indicator/item missing for measuring this dimension?

3. 2. 4. 4. Results A qualitative analysis was made of the information gathered in the discussion group sessions. This analysis led to the following conclusions. 1. None of the items was considered unsuitable by the collectives involved in the focus groups. All of them, without exception, were deemed to be appropriate, important and observable. 2. Several items were redrafted or qualified. Indeed, it might be said that the focus groups’ main contribution was to redraft several items or add clarifications or notes to some of the items on the scale. These redrafts or clarifications were in all cases designed to render the item’s content more specific and ensure it was applicable to everyone.

3. A very small number of items were proposed (N = 11) and none of them was specific to one collective, but rather applicable to any user of social services. This meant maintaining the 64 items that made up the initial version of the GENCAT Scale and including five new ones proposed by the focus groups and an indicator.

3. 2. 5. Pilot version of the GENCAT Scale The outcome of the process described above was the pilot version of the GENCAT Scale. The instrument finally consisted of 69 items – the 64 included in the initial version plus the five selected by the focus groups. This was the Spanish language version of the GENCAT Scale (Verdugo, Arias, Gómez & Schalock, 2009).

3. 2. 6. Translation of the GENCAT Scale into Catalan Given that application was to be made in Catalonia, the next logical step was to translate the pilot version of the scale into Catalan. The entire process of developing and translating the GENCAT Scale upheld the guidelines for adaptation and translation proposed both by the International Tests Commission (ITC) (Bartram, 2001; Hambleton, 1993, 1994, 1996; Muñiz & Hambleton, 1996; van de Vyjver & Hambleton, 1996) and by the International Quality of Life Assessment Project (IQOLA) (Aaronson et al., 1992).

3. 2. 6. 1. Translation into Catalan Two translations of the instrument were made from Spanish into Catalan. Both translations were carried out by two bilingual translators, whose mother tongue was Catalan. The translations were compared to single out any discrepancies between them and highlight the most ambiguous terms. Working with the original questionnaire at the same time as with the versions produced by the first (T-A) and second (T-B) translators, a synthesis was made to produce a common translation (T-AB).

3. 2. 6. 2. Back-translation Based on this T-AB version (synthesis of the two versions translated into Catalan), two members of INICO made two separate translations back into Spanish (BT-A and BT-B). On this occasion, Spanish was the two translators' mother tongue, although they were fluent in Catalan. Neither of them was aware of the concepts being investigated nor familiar with the original version.

3. 2. 6. 3. Committee of experts The committee of experts was made up of a methodologist, two experts in quality of life, two language professionals and the two back-translators. In addition, close contact was maintained with the two translators into Catalan throughout the entire process. The role of the committee of experts was to prepare the translated version for the field work. This involved reviewing all the translations and back-translations. Following a discussion process, consensus was reached on any discrepancies and equivalence was guaranteed between the original questionnaire and the translated version (semantic, linguistic and conceptual parity). No significant discrepancy was found between the versions, whereby the synthesis of the two versions translated into Catalan (T-12), with no modification, was the version used in the field work.

3. 2. 6. 4. Testing the translated version The translated version was applied to a further two members of INICO in order to ensure that it maintained equivalence with the original version in an applied situation. In view of the fact that no difficulty arose in the application procedure, the scale’s translation process was adjudged to have been completed. Consequently, now with the pilot version of the GENCAT Scale and its translation into Catalan, the process for developing and building the instrument was also deemed to have concluded. At this point, we should mention that the rigorous and thorough process followed has served and currently serves as a model for the development of other multidimensional quality of life instruments (Verdugo et al., 2007).

CHAPTER 4. VALIDATION BY MEANS OF CLASSICAL TEST THEORY (CTT) AND CONFIRMATORY FACTOR ANALYSIS (CFA)

4. 1. Goal The overriding aim of this study was to validate the GENCAT Scale by means of classical quantitative methods derived from CTT and using CFA. The general aim was broken down into three specific goals: 1.

Prove the scale’s validity through: (a) evidence based on the test’s content; and (b) evidence based on its internal structure.

Prove the scale’s reliability. This will involve an evaluation of: (a) its internal consistency; and (b) the standard error of measurement.

Prove the reliability of the items through: (a) index of reliability; and (b) index of difficulty.

The decision to prove its validity before proving its reliability was informed by recent work by Batista-Foguet, Coenders & Alonso (2004), as well as by Raykov & Marcoulides (2008).

4. 2. Method 4. 2. 1. Sampling procedure Involving the recipients of social services provided by the Catalan Welfare and Social Services Institute (‘Instituto Catalán de Asistencia y Servicios Sociales’) (ICASS), the selection of the sample was undertaken by means of a stratified and polytypic probability sampling. Stratification was performed according to both the nature of the target collective for the services and the geographical area of the centre providing such services. The sampling design described was applied to two differentiated groups. The first case involved a sampling of the population of elderly people (in rest homes and day centres). The second involved a sampling of the population of other collectives at risk of social exclusion: people with intellectual disability, physical disability, mental health problems, problems of substance abuse and those with HIV or AIDS.

4. 2. 1. 1. Sampling of elderly people The sampling of elderly people in rest homes and day centres was based on an estimation that 1,110 elderly people should be part of our sample and that the error should not exceed 3%. The sample was made with uniform allocation of the sample (90 sampling units for each geographical area or stratum) and with proportional allocation (a proportional number that depended on the total number of centres for each collective in the different geographical spheres). This meant that the sampling error was 2.91%. However, when considering solely rest homes the error rose to 3.62%, and in the case of day centres the sampling error was 4.74%. Nevertheless, the number of participants making up the sample observed was significantly higher (N = 1,619). For this reason, the sampling error for the total of rest homes and day centres in Catalonia was reduced to 2.43%.

4. 2. 1. 2. Sampling of the remaining collectives Regarding the sampling of all the other collectives, we were unaware of the size of the total population, but it was decided that the sampling error should not exceed 3% and we set the sampling size at 1,100 people. Here, too, application was made of a stratified probability sampling with a uniform allocation of the sample (105 sampling units for each geographical area or stratum) and with proportional allocation (a proportional number that depended on the total number of centres for each collective in the different geographical spheres). The sampling error for all the people at risk of exclusion was estimated to be 2.91%. Nevertheless, the sampling size increased to N = 1,410 and, therefore, the sampling error fell to 2.62%. The proportion of centres selected in each geographical area was undertaken by simple random sampling and the choice of the specific people to be evaluated in each centre by casual, incidental or convenience sampling.

4. 2. 2. Participants Our study has involved 608 professional staff from 239 entities providing social services and attached to the ICASS. Their task consisted in completing a scale for the objective evaluation of quality of life for 3,029 users of the aforementioned services in the Autonomous Community of Catalonia. In the 239 centres, the mean number of users evaluated was 12.67 (DT = 7.75) and the mean number of persons evaluated by each professional was approximately five (M = 4.981). Regarding those social services provided by the ICASS, the study involved those that provided services to the elderly (rest homes and day centres), people with intellectual disability, physical disability, mental health problems, problems of substance abuse and those with HIV or AIDS. The requirements laid down for allowing the professional staff in the aforementioned centres to take part in the study by completing the objective scales on quality of life were as follows: (a) to be familiar with the concept of quality of life and its evaluation, as well as be able to adopt the role of external observer and report objectively; (b) to work directly with the user and know the person well (to have known them for at

least three months); (c) to be able to observe the person in different contexts and during ample periods of time (at least several hours in each context). Regarding the users of the social services evaluated, the sole criteria governing their inclusion in the research were: (a) to be users of some kind of social services attached to the ICASS (and have done so for at least three months); (b) to be over the age of 18.

4. 2. 3. Instrument The instrument used was the GENCAT Scale: an instrument for the objective evaluation of quality of life designed according to the advances made on the eightdimension model proposed by Schalock & Verdugo (2002). The GENCAT Scale evaluates a series of observable aspects related to the eight areas that make up a person's quality of life and which may be the focus of customised support programmes arranged by different types of social services. The validation process used the Catalan language version (Verdugo, Arias, Gómez & Schalock, 2008a). The GENCAT Scale is used for the objective evaluation of the quality of life of adult users of social services. The ones completing the scale are the professional staff of these services who know well the person whose quality of life is to be evaluated and who have had recent opportunities to observe that person over prolonged periods of time and in different facets of their life. In total, the questionnaire has 69 items that cover observable aspects related to quality of life. They were all drafted as statements using the third-person and were randomly arranged around the eight subscales that correspond to the dimensions of the underlying theoretical model. Approximately half the items have a positive valency (n = 35) whilst the other half have a negative one (n = 34). The format for answering involves four frequency options: ‘never or hardly ever’, ‘sometimes’, ‘often’, ‘always or almost always’. Nevertheless, regarding those items that might be difficult to rate with this frequency scale, instructions have been given to answer bearing in mind a four-point Likert scale: ‘strongly agree’, ‘agree’, ‘disagree’, ‘strongly disagree’.

Although all the items are observable, specific and easy to understand, below each subscale a few simple clarifications were added for some of them. The items are preceded by brief instructions and two sociodemographic data forms. The first one is used to collect the details of the person being evaluated and the second one corresponds to the details of the professional completing the scale. The latter also includes the details of other people that might have provided the information in the event that the professional did not know the answer to any one of the items. For its correction, different yardsticks were provided for users of social services in general and specific ones for the various collectives: the elderly (aged over 50); people with intellectual disability; and all the other collectives at risk of social exclusion (people with physical disability, problems of mental health, substance abuse and HIV/AIDS). The composite scores recorded in each dimension and for the overall scale are converted according to these yardsticks into standard scores (M = 10; DT = 3), into percentiles and into a Quality of Life Index (M = 100; DT = 15). These scores allow identifying a person’s Quality of Life Profile in order to draw up person-centered support programmes and provide a reliable measure for monitoring the programme’s progress and results.

4. 2. 4. Application procedure Once a selection had been made of the centres taking part in the process for validating the GENCAT Scale (N = 351) we applied a strict application protocol. Firstly, the Catalan Welfare and Social Services Institute (ICASS) sent them an official letter explaining the purpose of the study and inviting the centres and services to take part. This letter was sent by both standard post and e-mail in order to ensure it was received at least in one of the ways. The research team then phoned each one of the centres selected with a view to: (a) confirming it had received the letter from ICASS; (b) confirming its postal address for sending it the scales via an express courier service; (c) informing it in greater detail about the research project and the number of users it was to evaluate; (d) confirming its involvement in the research; and (e) being at its disposal by phone or e-mail for any queries, comments or suggestions it might have regarding the scale.

Once the participation of all the centres included in the sample had been confirmed, the next step involved using an express courier service to send each one a package containing: (a) a manual with the necessary instructions for completing the scale; and (b) the specific number of scales to be filled in by each centre plus five more to ensure reaching the total number of participants considered ideal for validating the scale (N = 2,210). In total, 4,500 scales were sent out and 3,029 were returned (67.31%). At this point, we should like to stress the vital importance contact by phone had regarding our study's high response rate. Although calling by phone requires much more time and effort than other methods, we positively recommend it as it increases the chances of a high response rate.

4. 3. Results In order to conduct the analyses presented in this section, we used the statistical package SPSS 15.0 (SPPS, 2006), SAS 9.1.3 (SAS Institute, 2007), LISREL 8.8 (Scientific Software International, 2006) and Statistica 8.0 (StatSoft, 2007) for Windows.

4. 3. 1. Validity Different corroborations of the classical understanding of the GENCAT Scale’ validity are to be found in Verdugo, Arias, Gómez & Schalock (2008b). In order to imbue this work with a greater degree of innovation, we have opted to align ourselves with new conceptions (Messick, 1975, 1980, 1988, 1989, 1995, 1996, 1998; Popham, 2000; Zumbo, 2007), whereby it is affirmed that validity: (a) is a unitary concept based on different types of evidence; (b) implies a process, it is not an action that takes place in a single moment; (c) is a property of the scores, not a property of the instrument as such; (d) is a question of degree, not of all or nothing, whereby one should avoid referring to results as valid or non-valid; e) is always specific to a particular use or interpretation of the scores; and (f) involves an overall evaluative judgement in terms of support that guarantees its interpretations.

4. 3. 1. 1. Validity evidence based on scale content

Four procedures were used to corroborate this validity: a review of the scientific literature, the concordance between expert referees, the validation and redrafting of the items by the focus groups and, finally, an analysis of the discrimination power of the items. Given that the first three have been addressed in the preceding section, we shall use this section to present and interpret the results obtained in the analysis of the discrimination power of the items. Our aim when analysing the discrimination power of the items was to find out whether they allowed discriminating between previously established groups (those arising from the use of C25 and C75 as cut-off point) in each one of the eight dimensions and for the overall score of the GENCAT Scale. These three groups were called: people with low, medium and high scores. Given the ordinal nature of the data, we compared them with the Kruskal-Wallis nonparametric test. Before going into detail on the results obtained for each one of the dimensions and the total for the scale, we should note that all the correlations were significant (p = .000) in each and every one of the items. Accordingly, given the results forthcoming, we can conclude that the items on the GENCAT Scale discriminate significantly between the three groups specified, both when we take all the items together and when we analyse them by subscales. Nevertheless, we should remember that the high number of participants has a bearing on the test’s level of significance.

4. 3. 1. 2. Validity evidence based on internal structure

An analysis of a test's internal structure allows highlighting the extent to which the relationships between the items correspond to the construct being evaluated. Confirmatory factor analysis (CFA) is one of the techniques most widely used accordingly when a researcher has hypotheses on the structure of the latent variables, their inter-relationships and their relationships with the variables observed (conceptual model) (Batista & Coenders, 2000; Bollen, 1989; Byrne, 1998; Kaplan, 2000; Kline, 2005; Loehlin, 2004; Marcoulides & Schumacker, 2001).

a) Preparing the data Firstly, the data were screened and we noted the absence of univariate and multivariate normality, the low percentage of lost cases (random and no higher than 2.20% in any item), the nature of the outliers, the scarce linearity of the data and the absence of multicollinearity. Given these conditions and the high number of items, all the CFAs were undertaken on 32 parcels by means of the DWLS (Diagonal Weighted Least Squares) estimation method and on the matrix of polychoric variances-covariances and the estimation of asymptotic covariances.

b) Specification and identification of models For the formal definition of the model’s structure we based ourselves on a theoretical grounding. Accordingly, although theory clearly tends toward an eight-factor correlated model (Schalock & Verdugo, 2002), with a view to achieving model parsimony we tested the unidimensional structure. In addition, we checked a third model proposed in recent research (Wang et al., in press), in which quality of life is understood to be a hierarchical structure in which the eight dimensions are drawn together in a second-order one (quality of life). There now follows a further comparison of hierarchical structures by verifying the data fit to different first-order combinations of eight dimensions grouped into three second-order ones. These designs are reported in the paper we have just cited by Wang et al. (in press) and involve the so-called Schalock model and Salamanca model. All the models tested are listed and described forthwith.

Model 1:

Quality of life is a unidimensional construct.

Model 2:

Quality of life consists of 8 correlated factors (Model by Schalock & Verdugo, 2002): Self-determination, Social inclusion, Interpersonal relations, Rights, Material wellbeing, Emotional wellbeing, Physical wellbeing and Personal development.

Model 3:

Quality of life consists of 8 first-order factors and 1 second-order one (quality of life) (Model by Wang et al., in press).

Model 4:

Quality of life consists of 8 first-order factors and 3 second-order ones (Salamanca Model): ‘Personal Wellbeing’, ‘Empowerment’ and ‘Physical and Material Wellbeing’ (Wang et al., in press).

Model 5:

Quality of life consists of 8 first-order factors and 3 second-order ones (Schalock Model): Independence, Social Integration and Personal Wellbeing (Wang et al., in press).

c) Estimating the models' parameters In the first solution, corresponding to the unidimensional model (Figure 1), although all the coefficients were significant (‘t-values’ significantly different from zero) we noticed that the prediction error (θ ) ranged between .36 (p6_2) and .96 (p3_1). It is thereby deduced that the squared coefficient of multiple correlation (R2) for each indicator fell within a range of between .04 and .64. Moreover, only four of these coefficients are above . 50. This means that the proportion of variance in the variables observed that can be explained by the latent factor (quality of life) is far from suitable. In turn, the factorial saturations (λ) range between .27 (p3_1) and .80 (p6_2), with 14 out of the 32 factor loadings being lower than .50. In short, a single dimension is not enough to reproduce the original covariance matrix, and so we tested a multifactorial design.

e11 e12 e13 e14 e21 e22

.60 .55 .54 .56 .64 .93

e23

.64

e24

.70

e31

.96

e32

.87

e33

.93

e34

.96

P1_1 P1_2 P1_3

.63

.29

.67

.46

.68

.27

P5_1

.91

e51

P5_2

.79

e52

P5_3

.93

e53

.93

e54

.48

e61

.36

e62

.46

e63

.44

e64

.83

e71

.81

e72

.69

e73

.84

e74

P1_4

.66

.27

P5_4

P2_1

.60

.72

P6_1

P2_2

.26

.80

P6_2

P2_3

.60

.73

P6_3

P2_4

.55

P3_1 P3_2

QOL

.21 .37

.75

.41

P6_4 P7_1

.44

P7_2

.56

P7_3

.41

P7_4

P3_3

.27

P3_4

.20

e41

.67

P4_1

.57

.51

P8_1

.74

e81

e42

.51

P4_2

.70

.64

P8_2

.59

e82

e43

.51

p4_3

.70

.41

P8_3

.83

e83

e44

.60

.63

.49

P8_4

.76

e84

P4_4

Figure 1. Standardised parameters for model 1 (unidimensional)

The second model tested was the one by Schalock & Verdugo (2002): quality of life comprises eight dimensions that correlate with each other. The factorial design of this model (Figure 2) shows that the prediction errors (θ) range between .20 (p6_2) and .86 (p2_2). Therefore, the coefficients of determination (R2) (Table 1) range between .14 and .80. Half these values are above .50. Regarding factor loadings (λ), they fall within the range of between .38 (p2_2) and .89 (p6_2). 81.3% of the factorial saturations exceed the value of .50 and they were all statistically significant with t values of more than 2.58 (p< .01). Furthermore, all the coefficients were significant.

θδ1

.39

θδ2

.34

θδ3

.33

θδ4

.34

θδ5

.44

θδ6

.86

θδ7

.49

θδ8

.55

θδ9

.73

θδ10

.31

θδ11

.52

θδ12

.76

θδ13

.59

θδ14

.40

θδ15

.40

θδ16

.51

P1_1 P1_2 P1_3

.78 .63

.81

.40

.82

.50

P1_4

P2_2

.38

P2_3

P3_2 P3_3

.34

.71

.46

θδ18

P5_3

.77

θδ19

P5_4

.82

θδ20

P6_1

.32

θδ21

P6_2

.20

θδ22

P6_3

.30

θδ23

P6_4

.28

θδ24

.51

P7_1

.74

θδ25

.50

P7_2

.75

θδ26

P7_3

.57

θδ27

P7_4

.79

θδ28

P8_1

.55

θδ29

P8_2

.28

θδ30

P8_3

.70

θδ31

P8_4

.55

θδ32

.85

.31 .07

.53

.49

.54

.82

.59

.75

.67

.77

.70

.85 .54 .67

.64

P4_4

.66 .46

.38

.25

.77

.84

.37

.70

.69

.89

.56 .68

.83

P4_2

SD .66

P3_4

p4_3

.31

.35

.78

.24

.93

.52

.64

P5_2

.48

.83

.67

P4_1

θδ17

.73

.29 63

P2_4 P3_1

.76

.60

.77

P2_1

P5_1

.43

.81

.75

.49

Note. EM = Emtional weelbeing; IR = Interpersonal relations; MW = Material wellbeing; PD = Personal development; PW = Physical wrllbeing; SD = Self-determination; SI = Social inclusion; and RI = Rights

Figure 2. Standardised parameters for model 2 (Schalock & Verdugo, 2002)

Table 1. Factor loadings (λ), prediction errors (θ) and coefficients of determination (R2) for model 2 Emotional Well-being Parce

Interp. Relations Parce

Material Well-being R2

Parce

Personal Development Parce

p1_1

.78

.39

.61

p2_1

.75

.44

.56

p3_1

.52

.73

.27

p4_1

.64

.50

p1_2

.81

.34

.66

p2_2

.38

.86

.14

p3_2

.83

.31

.69

p4_2

.78

.40

.60

p1_3

.82

.33

.67

p2_3

.71

.49

.51

p3_3

.69

.52

.48

p4_3

.77

.40

.60

p1_4

.81

.34

.64

p2_4

.67

.55

.45

p_34

.49

.76

.24

p4_4

.70

.51

.49

Physical Well-being Parce

Self-Determination Parce

Social Inclusion Parce

Rights R2

Parce

p5_1

.49

.76

.24

p6_1

.83

.32

.68

p7_1

.51

.74

.26

p8_1

.67

.55

.45

p5_2

.73

.46

.64

p6_2

.89

.20

.80

p7_2

.50

.75

.25

p8_2

.85

.28

p5_3

.48

.77

.23

p6_3

.84

.30

.70

p7_3

.66

.57

.43

p8_3

.54

.70

.30

p5_4

.43

.82

.18

p6_4

.85

.28

.72

p7_4

.46

.79

.21

p8_4

.67

.55

.45

Insofar as the third model is concerned, proposed by Wang et al. (in press), in which quality of life is understood to be a hierarchical structure in which the eight basic dimensions are grouped into a single higher-order factor (quality of life), we found that the data did not fit the model. So much so, in fact, that the software program could not reach a standardised solution and, therefore, it was not possible to obtain standardised parameters (Figure 3). It was not even possible to obtain them by means of other much less restrictive estimation methods such as ULS (‘Unweighted Least Square’). Accordingly, it can be concluded that the theoretical model in no way fits the empirical data in our study, and therefore does not constitute a plausible approach to the empirical data used.

e11 e12 e13 e14 e21 e22

.60

P1_1

.55

P1_2

.54

ζ5 PW

P1_3

.56

P1_4

.64

P2_1

.93

ζ1

ζ2

P2_2

e23

.64

e24

.70

P2_4

e31

.96

P3_1

e32

.87

P3_2

e33

.93

e34

.96

P3_4

e41

.67

P4_1

e42

.51

P4_2

ζ6 IR

P2_3

p4_3

e44

.60

P4_4

e51

P5_2

.79

e52

P5_3

.93

e53

P5_4

.93

e54

.48

e61

.36

e62

.46

e63

.44

e64

.83

e71

.81

e72

.69

e73

.84

e74

P8_1

.74

e81

P8_2

.59

e82

P8_3

.83

e83

P8_4

.76

e84

P6_1 P6_2

P6_4

QOL

ζ3

P3_3

e43

.91

P6_3

.51

P5_1

ζ7

P7_1 P7_2

P7_3 P7_4

ζ4

ζ8 PD

Figure 3. Standardised parameters for model 3 (Wang et al., in press)

An initial examination of the fourth model, the so-called Salamanca model (Figure 4), reveals several offending estimates. Thus, saturation λ of the endogenous latent variable Physical wellbeing and the correlation Φ between the higher-order factors Empowerment and Personal wellbeing adopt values higher than 1. Along these same lines, a variance of zero (ζFW) is obtained. Accordingly, although the model has been theoretically identified, there are problems regarding empirical identification.

θε1 θε2 θε3 θε4

.40 .34 .32 .33

θε5

.46

θε6

.89

θε7

.46

θε8

.54

θε13

.59

θε14

.40

θε15

.40

θε16

.51

θε21

.30

θε22

.20

θε23

.31

θε24

.28

ζ1 P1_1

.49

.27

ζ7

.77

P1_2 P1_3

.49

.81

.51

.83

.66

.85

.82

ζ2

.71

P2_3

.33

ζ8 .51

.86

.74

.55

.72

.82

P8_1 .67

.70

1.05

.68

Personal Wellbeing

ζ4

.56

P4_4

.73

ζ6

.46

.83

.83 .85

P6_4

.48

1.00

P5_2 P5_3

.42

P5_4

Physical & Material Wellbeing

ζ3 .60

P3_1 .52

.63

.74

.70

P6_3

P5_1 .49

.78

.88

ζ5

.93

.78

P6_1

.00

.13

.64

P6_2

P8_3 P8_4

P4_1

p4_3

P8_2

.65

P2_4

P4_2

P7_3 P7_4

Empower ment

.33

.74

P2_2

P7_2

.47

P1_4

P2_1

P7_1

.82 .69

P3_2 P3_3

.50

P4_3

.76

θε25

.74

θε26

.57

θε27

.78

θε28

.55

θε29

.25

θε30

.70

θε31

.58

θε32

.76 .45

θε17 θε18

.77

θε19

.83

θε20

.73

θε9

.33

θε10

.52

θε11

.75

θε12

Figure 4. Standardised parameters for model 4 (Salamanca model)

θε13

.42

θε14

.19

θε15

.23

θε16

.33

θε21

.19

θε22

.15

θε23

.25

θε24

.22

θε5

.20

θε6

.66

θε7

.36

θε8

.43

θε25

.45

θε26

.48

θε27

.36

θε28

.64

θε29

.39

θε30

.32

θε31

.58

θε32

.56

ζ4

P4_1

.06

.76

P4_2 p4_3

.90

.88

.97

.82

P4_4

.07

ζ6

P6_1

.13 .90

P6_2 P6_3

Independ ence

ζ1 P1_1 .88 .89

.89

.93

.92

P1_2 P1_3

.88

P1_4

.97

.87

.22

θε1

.20

θε2

.21

θε3

.23

θε4

.70

θε17

.96

.88

P6_4

ζ2

P2_1

.97

Personal wellbeing

.06

.90

P2_2 P2_3

.98

.58

.30

.75

.55 .85

.90

ζ7

P7_1

P5_1

.74

.80

P2_4

ζ5

.53

P7_3

.72

P5_4

Social integration

P7_4

ζ8 .78

P8_2 P8_3

.83 .65

ζ3

.94

.60

P8_1

.11 .94

θε18

.72

θε19

.74

θε20

.80

θε9

.28

θε10

.54

θε11

.77

θε12

.84

.80

P5_3

.51

.11

.74

P7_2

P5_2

.28

.45

P3_1 .45

.85 .68

P3_2 P3_3

.48

P4_3

.67

P8_4

Figure 5. Standardised parameters for model 5 (Schalock model)

Finally, the last model subjected to confirmation was the so-called Schalock model. On this occasion, the standardised solution (Figure 5) gave rise to prediction errors (θ) of between .15 and .80 for the parcels p6_2 and p3_1, respectively (therefore, the coefficients of determination range between .20 and .85). Regarding the factor loadings of the endogenous variables on the variables observed (λ), there are fairly high values that range between .45 and .92. The same circumstance, albeit much more pronounced (i.e., excessively high values), occurs in the factor loading of the endogenous variables on the exogenous ones (.74 ≤ γ ≤ .97) and in the correlations between the exogenous variables (Φ ≥ .90).

d) Goodness of fit of the models Once the parameters have been estimated, the final step in the CFA consists in evaluating the fitting of the theoretical models to the study data. Table 2 therefore presents some of the more common goodness of fit indices. The overall or absolute fit index traditionally used for verifying the null hypothesis (i.e., the model fits the population data perfectly) is the Satorra-Bentler Chi-Square Index (Satorra & Bentler, 1994). When analysing the values returned by all the models, we would have to reject the null hypothesis in all cases (p = .000). Nonetheless, from a more pragmatic and less restrictive perspective, it is advisable to examine not so much the level of statistical significance but rather the magnitude of χ2 (Arias, 2008): high values would correspond to a deficient fit and low values to a better fit. Therefore, the second model (i.e., quality of life is a construct made up of eight inter-related factors) is by far the one recording the lowest value, (χ2S-B = 1251.16); although all the values returned are fairly high given that this index is strongly influenced by the sample size. Concerning the ratio of χ2/gl, the second model (Schalock & Verdugo, 2002) is the only one attaining an acceptable value.

Table 2. Fit indices for the various models Fit indices

Model 1

Model 2

Model 3

Model 4

Model 5

5974.82

1251.16

5572.16

1792.28

1779.39

464

436

456

453

.000

12.88

2.87

12.22

3.96

3.93

SRMR

.13

.076

.18

.097

.096

GFI

.89

.96

.77

.94

AGFI

.78

.96

.73

.93

.92

Parsimoniou

RMSEA

.15

.058

.19

.073

Pclose

.000

Incremental

NFI

.76

.95

.77

.93

TLI

.75

.96

.77

.94

χ2S-B

Absolute

χ2/gl Partial

Absolute

CFI

.77

.97

.79

.94

IFI

.77

.97

.79

.94

RFI

.74

.94

.75

.92

Precisely for the above reason and with a view to obtaining a more reliable representation of the true goodness of fit of the models, it is advisable to consider another type of indices (Cea, 2002; Roussel et al., 2002). In order to overcome the drawbacks posed by the overall fit index, a raft of partial fit indices have been developed. Amongst these are the fit indices of an absolute, parsimonious and incremental nature. Accordingly, the partial fit indices of an absolute nature GFI (Goodness-of-Fit Index) and AGFI (Adjusted Goodness-of-Fit Index) assess the degree to which the model’s variances and covariances correctly reproduce the original matrices (the former does so in an absolute manner whilst the latter do so in keeping with the degrees of freedom). Both should exceed the value of .90; a condition fulfilled only by the second, fourth and fifth models, albeit the second with greater excellence (.96). The SRMR (Standardized Root Mean Square), for its part, records no more than an acceptable value in the case of the second model. Regarding the parsimonious fit indices, only the second model records a value of RMSEA (Root Mean Square Error of Approximation) that can be considered good (RMSEA = .058). The fourth and fifth models, however, record reasonable values (RMSEA = .073), whereby the unidimensional model and the one considering a first-order factor attain values that lead to the rejection of both models (above .10). On the other hand, the probability of the RMSEA value being lower than .05 is < 1% in all cases (P-Value for Test of Close Fit = .000). Finally, the incremental fit indices: NFI (Normed Fit Index), TLI (Tucker-Lewis Index), CFI (Comparative Fit Index), IFI (Incremental Fit Index) and RFI (Relative Fit Index), assess the extent to which one model is better than the others. According to these, the first and third model should be rejected (in fact, it should be remembered that the third model did not reach a standardised solution, which makes it somewhat meaningless to interpret the fit indices) by recording values ≈ .70. In the rest of the models, the values show a good fit (≈ . 90), although the second model is the one returning the manifestly best coefficients (.94 - . 97). Consequently, the results obtained clearly show that the second model (Schalock & Verdugo, 2002), in which quality of life is understood to consist of eight basic inter-related dimensions, is by far the one that best fits the data and, therefore, the one that will be used for the subsequent analyses.

e) Reliability and validity of the model by Schalock & Verdugo Regarding the indices of reliability and validity of the final model (model 2: Schalock & Verdugo, 2002), and in addition to the reliability of the individual indicators (see R2 in Table 1), computation was made of the composite reliability of each latent variable (i.e., the internal consistency of the eight constructs or reliability of the constructs) and of the model, and the average variance extracted for each one of the latent constructs (i.e., validity or degree to which the indicators accurately measure the corresponding construct) and for the model. All the values mentioned are included in Table 3. Table 3. Composite reliability (ρc) and average variance extracted (ρv)

ρc

ρv

Emotional wellbeing

.881

.648

Interpersonal relations

.729

.415

Material wellbeing

.734

.419

Personal development

.815

.525

Physical wellbeing

.618

.297

Self-determination

.914

.727

Social inclusion

.614

.289

Rights

.782

.478

Model

.965

.938

As can be seen in Table 3, the reliabilities of four of the listed constructs exceed the threshold of .75; whist the average variance extracted is over 50% in three cases. This means that the values obtained in Self-determination, Emotional wellbeing, Personal development and Rights reflect positively on the validity and reliability of the indicators used for the empirical explanation of the latent constructs. The dimensions that come out worse in the model are Material wellbeing, Interpersonal relations, Physical wellbeing and, especially, Social inclusion. Regarding the model’s indicators of reliability and validity, the indices obtained were more than satisfactory (> .93). It is therefore concluded that the eight-dimension model considered by Schalock & Verdugo (2002) presents a suitable portrayal of the quality of life of the users of social

services provided in Catalonia by ICASS, which is validity evidence based on the factor structure of the GENCAT Scale. f) Relations between latent variables in the model by Schalock &Verdugo Table 4 shows the coefficients of correlation between the model’s latent variables. Amongst them, there are two that are scarcely satisfactory: by excess, the one observed between the dimensions of Interpersonal relations and Social inclusion (r = .93) and, at the other end of the scale, the one recorded between Material wellbeing and Self-determination (r = .07). Most of the correlations, however, can be deemed acceptable (they range between . 24 and .82) considering the content referred to by each pair of latent constructs. Therefore, with the exception of those noted, we consider that the correlations listed provide further evidence of the scale’s validity, inasmuch as they are consistent with the results of prior research.

Table 4. Coefficients of correlation between the model’s latent variables EW

.77

.34

.35

.68

.70

.25

.63

.50

.63

.54

.40

.31

.07

.75

.24

.60

.93

.53

.77

.66

.56

.29

.37

.38

.64

.31

.82

.59

4. 3. 2. Reliability of the scale

Given that we have only a single application of the GENCAT Scale, its reliability was verified in terms of internal consistency and standard error of measurement.

Furthermore, we have assessed the reliability of the items by means of the indices of reliability and difficulty.

4. 4. 2. 1. Internal consistency and standard error of measurement The GENCAT Scale has an excellent Cronbach’s Alpha coefficient (α = .916), a reliability coefficient that is only partially satisfactory (r = .786), acceptable lambda coefficients (λ1,2,3 > .900), satisfactory coefficients arising from the covariance of the items ( θ = Ω = .925) and an acceptable error of measurement (TEM = 6.923). The analysis by subscales revealed that the dimensions Self-determination (α = .880), Emotional wellbeing (α = .834) and Personal development (α = .737) were the most reliable, whereas Physical wellbeing (α = .474), Material wellbeing (α = .569) and Social inclusion (α = .570) were the least reliable ones. Interpersonal relations and Rights were moderately reliable (α = .663 and .694, respectively). An analysis of the scale’s internal consistency overall and by dimensions reveals significant differences between the values obtained depending on the collectives involved in each case, except in the dimensions of Physical wellbeing and Social inclusion. Thus, the GENCAT Scale appears to have coefficients of internal consistency that are significantly higher than all the other groups for people with substance abuse in the dimensions of Emotional wellbeing, Interpersonal relations, Material wellbeing and the Total scale, whereas the coefficients that were significantly higher in the dimensions of Personal development and Rights are returned by the elderly. By contrast, the coefficients recorded for people with HIV/AIDS are significantly lower in three of the eight dimensions.

4. 4. 3. Reliability of the items An analysis of the reliability of the items for the overall scale reveals that 69, that is, 84.05% of the items can be considered suitable, although 11 of them record Corrected Homogeneity Indices (CHI) lower than .200 and five Reliability Indices (RI) lower than .100. In an analysis by subscales, the dimensions Self-determination and Emotional wellbeing contain the more reliable items, with a mean RI of .620 and .502, respectively. The dimensions that include the least reliable items are, in turn, Physical wellbeing (M = . 135) and Material wellbeing (M = .187). The items in the remaining dimensions record means ranging between .203 (Rights) and .399 (Personal development); although it should

be noted that Interpersonal relations (M = .220) includes an item with negative RI and CHI (item IR_09).

CHAPTER 5. VALIDATION OF THE SCALE USING MULTI-GROUP EXPLORATORY STRUCTURAL EQUATION MODELING AND CALIBRATION USING THE RASCH MODEL

5. 1. Goals and hypotheses The goals and hypotheses that will inform the study described in this chapter are those specified forthwith: Goal 1. Given that there are significant differences depending on the condition of the person evaluated, the first goal involved analysing whether there was equivalence in the factor structure of several groups according to Exploratory Structural Equation Modeling (ESEM) (Asparouhov & Muthén, in press). Hypothesis 1. The factor structure posited by Schalock & Verdugo (2002) in which quality of life consists of eight dimensions fits the data for elderly people (in rest homes and day centres) according to the ESEM model. Hypothesis 2. The factor structure posited by Schalock & Verdugo (2002) in which quality of life consists of eight dimensions fits the data for people in a situation of social disadvantage (people with intellectual disability, people with physical disability, people with mental health problems, people with substance abuse and people with HIV/AIDS) according to the ESEM model. Hypothesis 3. The factor structure posited by Schalock & Verdugo (2002) in which quality of life consists of eight dimensions is equivalent in the group of elderly people and in the group of people at social disadvantage according to the ESEM model.

Goal 2. Conduct a preliminary study of the instrument’s psychometric properties through the use of the Rating Scale Model (RSM) (Andrich, 1978; Wright & Masters, 1982). Goal 2.1.

Verify the unidimensionality of each factor.

Goal 2.2.

Obtain validity evidence for the instrument through the observed fit of the data to the model, regarding both items and persons.

Goal 2.3.

Calculate the indices of reliability and separation.

Goal 2.4.

Estimate the calibration of the items.

Goal 2.5.

Verify the accuracy of the measurement.

Goal 2.6.

Analyse invariance according to the person’s gender and condition.

Goal 2.7.

Determine whether there is Differential Item Functioning (DIF) by gender and type of collective.

5. 2. Method

5. 2. 1. Participants The theoretical sample was as described in the preceding chapter: 3,029 users of social services in Catalonia.

5. 2. 2. Instrument So, too, has the GENCAT Scale (Verdugo et al., 2008a) been described in Chapter 4 of this work.

5. 2. 3. Procedure The analyses based on ESEM have been conducted with MPlus v.5.2 (Muthén & Muthén, 2008), whereas the Rasch Rating Scale Model (RSM) (Andrich, 1978; Wright & Masters, 1982) has been implemented in the software WINSTEPS 3.68.0 (Linacre, 2008; Linacre & Wright, 1999).

5. 3. Results 5. 3. 1. Evaluation of factor invariance The comparison of factor equivalence between groups usually begins with a preliminary analysis in which an examination is made of the goodness of fit of a same model on each separate sample and group under study. The purpose of this preliminary analysis is to show that the basic model specified provides a good fit in both groups and is readily interpretable in each one of them. This analysis is based on the initial consideration of a model that has the same form for all the groups, although there is no restriction on the value that each parameter may have for each one of them following the estimation process (i.e., they have the same latent variables, the same observable variables and the relationships between them are identical, even though the value of their parameters may not be). Figure 6 presents the eight-dimension model posited by Schalock & Verdugo for each one of the groups. The performances of these analyses involved the use of both the GEOMIN rotation method, which is an oblique rotation with an epsilon value of .01 (Yates, 1987; Browne, 2001) and is recommended when the factor loadings are substantial in more than one factor, and the Weighted Least-Squares Method (WLSM), which is an estimation procedure for general analyses with categorical variables that uses the weighted diagonal matrix with standard errors and Chi-Square test with the complete weighted matrix. The fit of the eight-dimension model for both groups is shown in Table 5. Nonetheless, all the fit indices are slightly higher in the case of elderly people. Table 5. Fit indices for the eight-dimension model in elderly people and in those at social disadvantage (ESEM) Nature Fit indices Elderly Social disadvantage Absolute χ2 6589.607 8017.028 gl 1.822 1.822 p .000 .000 Scaling Correction Factor for MLR .444 .426 Partial Absolute SRMR .038 .040 Parsimoniou RMSEA .043 .051 s Incremental CFI .979 .965 TLI .973 .955

θδ1 θδ2 θδ3 θδ4 θδ5 θδ6 θδ7 θδ8 θδ9 θδ10 θδ11 θδ12 θδ13 θδ14 θδ15 θδ16 θδ17 θδ18 θδ19 θδ20 θδ21 θδ22 θδ23 θδ24 θδ25 θδ26 θδ27 θδ28 θδ29 θδ30 θδ31 θδ32 θδ33 θδ34

BE_1

BE_2

BE_3

BE_4

BE_5

BE_6

BE_7

BE_8

RI_1

RI_2

RI_3

RI_4

RI_5

RI_6

RI_7

RI_8

RI_9

RI_10

BM_1

BM_2

BM_3

BM_4

BM_5

BM_6

BM_7

BM_8

DP_1

DP_2

DP_3

DP_4

DP_5

DP_6

DP_7

DP_8

λ1,1 λ2,1 λ3,1 λ4,1 λ5,1 λ6,1 λ7,1 λ8,1 λ1,2 λ2,2 λ3,2 λ4,2 λ5,2 λ6,2 λ7,2 λ8,2 λ9,2 λ10,2 λ1,3 λ2,3 λ3,3 λ4,3 λ5,3 λ6,3 λ7,3 λ8,3 λ1,4 λ2,4 λ3,4 λ4,4 λ5,4 λ6,4 λ7,4 λ8,4

Φ15 Φ14 Φ13 Φ16 Φ17

ξ1

Φ12

ξ5

Φ58 Φ57 Φ56

Φ18

Φ25 Φ26

ξ2

Φ28

Φ24

Φ67

ξ6

Φ68 Φ23

Φ67

Φ27

Φ36

Φ37

ξ3

ξ7

Φ35

Φ38

Φ34

Φ45 ξ4

Φ78 Φ46 Φ47

ξ8

Φ48

λ1,5 λ2,5 λ3,5 λ4,5 λ5,5 λ6,5 λ7,5 λ8,5 λ1,6 λ2,6 λ3,6 λ4,6 λ5,6 λ6,6 λ7,6 λ8,6 λ9,6 λ1,7 λ2,7 λ3,7 λ4,7 λ5,7 λ6,7 λ7,7 λ1,8 λ2,8 λ3,8 λ4,8 λ5,8 λ6,8 λ7,8 λ8,8 λ9,8 λ10,8

BF_1

BF_2

BF_3

BF_4

BF_5

BF_6

BF_7

BF_8

AU_1

AU_2

AU_3

AU_4

AU_5

AU_6

AU_7

AU_8

AU_9

IS_1

IS_2

IS_3

IS_4

IS_5

IS_6

IS_7

IS_8

DE_1

DE_2

DE_3

DE_4

DE_5

DE_6

DE_7

DE_8

DE_9

DE_10

θδ35 θδ36 θδ37 θδ38 θδ39 θδ40 θδ41 θδ42 θδ43 θδ44 θδ45 θδ46 θδ47 θδ48 θδ49 θδ50 θδ51 θδ52 θδ53 θδ54 θδ55 θδ56 θδ57 θδ58 θδ59 θδ60 θδ61 θδ62 θδ63 θδ64 θδ65 θδ66 θδ67 θδ68 θδ69

Figure 6. Model for elderly people and for those at social disadvantage

As can be seen in Table 6, the data fit in the invariance analysis of the two groups with no kind of restriction provides partial fit indices that can be considered acceptable (Hu & Bentler, 1999); although they are not very high (CFI = .856; TLI = .814), they appear to provide a fairly approximate reflection of the degree of invariance between the two groups. We can therefore conclude that the model's form is upheld and is satisfactory for both groups. However, as suggested by the preliminary analysis of the groups made separately, the joint analysis revealed different contributions by these to Chi-Square.

Given both its ability to test hypotheses and its suitability for the analysis, the χ2 index is the one most used in the case of fitting multi-sample models. This, together with the calculation of the model’s degrees of freedom, will report on the probability of acceptance of the null hypothesis specified (Browne, 1982, 1984). This hypothesis maintains that groups are invariant, whereby if the probability associated to the χ2 statistic is greater than .05 (for an α = .95), it will confirm the equivalence of the parameters in the different groups. The interpretation of χ2 is therefore common to that of the analysis with a single group, as are all the other fit indices, although their calculation fits in with the simultaneous estimation of the parameters of several models (Marsh & Hocevar, 1994). Such contributions confirm a higher fit of the eight-dimension model in the group of elderly people in rest homes and day centres than amongst the people at risk of social exclusion. The testing of the factor invariance hypothesis for both groups was significant (p = .000) for a difference of χ2 = 988,912 and 292 degrees of freedom; that is, the saturations of the items in the factors work differently in each one, with the model’s fit being substantially better in the case of the former than in the latter.

Table 6. Fit indices for the multigroup invariance analysis without restrictions Multigroup invariance Nature Fit indices analysis 2 Absolute χ 13083.860 gl 3644 p .000 2 Contribution to χ of elderly people 6047.474 2 Contribution to χ of people at social disadvantage 7036.386 Partial Absolute SRMR .030 Parsimonious RMSEA .045 Incremental CFI .856 TLI .814 Log-likelihood Value of H0 -189400.533 Value H1 -182858.603 Information Number of free parameters 1.324 Akaike Information Criterion (AIC) 381449.066 Bayesian Information Criterion (BIC) 389185.802 BIC adjusted to the size of the sample 384979.097

5. 3. 2. Analysis with the Item Response Theory (IRT) Model Before describing the results obtained with the logistical analysis of a parameter (Rasch Model), more specifically with the Rating Scale Model (RSM) posited by Andrich (1978) for each one of the subscales or dimensions that make up the GENCAT Scale, we performed the pertinent prior verifications of the model’s data fit: (a) the point-biserial correlations were positive in all cases (.13 and .82); (b) the item-category empirical measurements observed appeared in the proper order according to the level of attribute in all the items except for three (‘Where he/she lives stops them from leading a healthy life’ in Material wellbeing, ‘He/she finds it difficult to access healthcare resources’ in Physical wellbeing and ‘The service he/she attends respects their privacy’ in Rights), in which categories 1 and 2 are swapped around; (c) the function of the categories was suitable in all cases; in fact, each category had more than 1,000 observations, comfortably surpassing the recommended minimum number of responses (i.e., 10 observations). Furthermore, the mean measurements for the categories (τ) progressively increased in all the dimensions except in Material wellbeing and Rights, in which category 3 was especially noisy; (d) the amount of variance explained by the measurements confirmed the unidimensionality of the eight subscales; and (e) the study of item misfit confirmed the sitting of the items within the recommended range (Linacre, 2002) with the sole exceptions being three items (‘He/she is exposed to exploitation, violence or abuse’ in Rights, ‘The service he/she attends caters for their preferences’ in Self-determination and ‘He/she has a satisfactory sex life’ in Interpersonal relations, which returned values slightly higher than 2). Once the prior verifications had been made for the model’s data fit, we proceeded to check the unidimensionality of each factor or subscale. Accordingly, the principal components analysis of the eight subscales’ residuals gave rise to percentages of between 36% and 58.7% of variance explained by the modelled data. More specifically, an analysis of the subscales Interpersonal relations and Self-determination gave rise to percentages that were slightly lower than the recommended value of 60%, whilst the majority exceeded the commonly used one of 40%. Only the subscales of Physical wellbeing and Material wellbeing did not reach that value, although they came very close. These results, together with the first tests in each principal components analysis (with eigenvalues below 3.0, which is considered to indicate the existence of a second dimension), led us to confirm the unidimensionality of the eight subscales of quality of life. Secondly, we analysed the model’s data fit. On the one hand, regarding the fit for people, it is noteworthy that Interpersonal relations is the only subscale that does not have extreme data,

whereas Rights and Material wellbeing have more than 600 extreme cases. In addition, we have a perfect fit for persons in Self-determination and Social inclusion in the case of both the infit and the outfit, and in Physical wellbeing in the case of the infit. All the other values ranged between -.1 and . 1. Finally, the values of MNSQ were very close to 1 in all cases. For these reasons, we can conclude that the overall fit for persons shows that the responses are consistent with the response patterns foreseen by the model. On the other hand, regarding the overall fit of the items, the MNSQ values confirmed the items’ fit to the RSM in all cases. On this occasion, Emotional wellbeing and Social inclusion were the subscales with a perfect fit and Material wellbeing exceeded the value |1.0|, albeit only slightly. Given that all the values fell within the range considered to be acceptable, we confirmed the overall fit of the items to the model. A more detailed analysis of the items’ fit furthermore revealed that the more accurate dimensions are Emotional wellbeing, Personal development, Self-determination and Social inclusion. By contrast, the least accurate items are ‘His/her workplace complies with rules on safety’ (MODEL S.E. = .05) and ‘Where he/she lives is clean’ (MODEL S.E. = .05) in Material wellbeing, ‘The service he/she attends supervises the medication they take’ (MODEL S.E. = .07) in Physical wellbeing, and ‘The service he/she attends upholds and defends their rights’ (MODEL S.E. = .05), ‘The service respects the privacy of information’ (MODEL S.E. = .05) and ‘He/she is exposed to exploitation, violence or abuse’ (MODEL S.E. = .08) in Rights. No item presented dependence or determinism (values below 0.60). However, five of the 69 items revealed a lack of fit, noise or high random variability in the data (values substantially higher than 1.5: ‘He/she has a satisfactory sex life’ in Interpersonal relations, ‘He/she finds it difficult to access healthcare resources’ and ‘The service he/she attends supervises the medication they take’ in Physical wellbeing, ‘The service he/she attends caters for their preferences’ in Self-determination and ‘He/she is exposed to exploitation, violence or abuse’ in Rights). In brief, most of the parameters for the items in the GENCAT Scale have an acceptable behaviour according to the postulates of the Rasch model, whereby they all have a suitable fit, with the exception of the final item (‘He/she is exposed to exploitation, violence or abuse’ in Rights, whose fit is highly debatable).

The calculation of reliability involved the use of the item separation reliability index and the person separation reliability index. Concerning the former, we obtained a separation reliability index for the items equal to 1 in all the subscales, so we can affirm that the items have the utmost reliability. However, the separation reliability indices for persons were lower and varied greatly between the subscales: Self-determination was the only dimension that returned a value considered to be acceptable (.82), whereas Material wellbeing returned a reliability coefficient that was so low (.15) that its reliability as regards persons is highly questionable. All the other values ranged between .23 (Physical wellbeing) and .77 (Emotional wellbeing). The results for the separation indices for items and persons confirmed the previous results, whereby the separation index for the items exceeds the value of 2.00 in all cases (in fact, it exceeded the value of 18 in all cases and reached a value of 39.19 in Interpersonal relations). However, the separation index for persons only exceeded the value of 2.00 in the case of the subscale Self-determination (2.16). The lowest separation indices corresponded to Material wellbeing (0.41) and Physical wellbeing (0.55). The next step involved the calibration of the items. Generally speaking, we can conclude that there is equilibrium in all the scales regarding the number of difficult items (above 0 logits) and easy items (below 0 logits). Regarding the distribution along the continuum of the dimension of quality of life they are evaluating, Interpersonal relations is the one with the greatest range and best distribution of items together with Social inclusion. On the other hand, the dimensions with the largest jumps or gaps between items are Material wellbeing (with a very difficult item in comparison to all the others: ‘He/she does not earn enough to be able to afford luxuries’), Personal development, Physical wellbeing, Self-determination and Rights (with a very easy item in comparison to all the others: ‘The service he/she attends takes their personal development into account’, ‘The service he/she attends supervises the medication they take’, ‘The service he/she attends caters for their preferences’ and ‘He/she is exposed to exploitation, violence or abuse’). Finally, Emotional wellbeing is the subscale whose items are grouped into a smaller range, albeit evenly distributed without large gaps between them. An analysis of the suitability of the items’ level of difficulty for the sample confirms these results, highlighting the almost perfect adaptation of the items in Interpersonal relations. For Emotional wellbeing, by contrast, it would be advisable to include not only easier items but also, and above all, more difficult ones. It is highly advisable to include more difficult items in Material wellbeing, Physical wellbeing and Rights, as most of those included are too easy for the participants.

In Personal development and Self-determination it would be convenient above all to include more difficult items, although there is a noticeable lack of one or two in the easier levels. Finally, in Social inclusion it would be appropriate to include more difficult items, whilst one item (‘He/she is rejected or discriminated against by others’) is too easy for the participants. Regarding the adaptation of the response categories to the sample, we observe that, although the extreme categories are the most probable ones in all cases, the four response options are suitable in all the subscales, except for Interpersonal relations, Material wellbeing and Rights, in which option 2 (‘sometimes’ or ‘often’ according to the item’s valency) is not very suitable. Furthermore, the information curves for the categories reveal that the extreme options provide the most information (‘never or hardly ever’ in first place, followed by ‘always or almost always’) in Interpersonal relations, Material wellbeing and Rights. Emotional wellbeing is the only subscale in which the intermediate options (‘sometimes’ or ‘often’) provide more information than the extreme ones. In Self-determination and Social inclusion, the category that provides the most information is ‘often’, followed by ‘never or hardly ever’. Finally, the category in Social inclusion and Physical wellbeing providing the most information is ‘never or hardly ever’, with the least being provided by ‘sometimes’. In order to evaluate the accuracy of the measurements, we analysed the information functions of the items and of the tests. In short, we can affirm that the items providing the least information are those in Emotional wellbeing (0.62), Personal development (0.86) and Selfdetermination (0.89). All the other dimensions exceed the value of 1.00 for information, with Rights recording the highest value (1.24). The information function of the tests also reveals that in Interpersonal relations, Material wellbeing and Rights, the tests provide the utmost information when

θ is between -0.5 and 0.0. Self-determination, Social inclusion, Physical wellbeing and

Personal development, in turn, provide their most information in the levels of the measurement falling between approximately -0.02 and + 0.02, whilst the region of maximum information in the subscale of Emotional wellbeing is situated between approximately -0.04 and 0.01.

Invariance was analysed by gender and by group or condition. The analysis of the invariance of the items between men and women gave rise to 12 items that fell outside the margin of error at a 95% confidence interval. However, an analysis of the DIF confirmed solely the differential function of one item: i39 ‘He/she maintains good personal hygiene’ in Physical wellbeing, which was much easier for women than for men. If we consider its content, we cannot affirm that there is a bias in favour of one gender or the other. In turn, the analysis of the invariance of groups (elderly people on the one hand and people at social disadvantage on the other) was much more contentious: 36 items were clearly located outside the area of invariance and a similar number were borderline, whereby their invariance was also highly debatable. Invariance was clear solely in the case of 14 items. Nonetheless, the analysis of the DIF confirmed only the differential function of 10 of them: ‘He/she has problems of conduct’ in Emotional wellbeing, which was more difficult for people in a situation of social disadvantage; ‘Most of the people with whom he/she interacts are in a similar situation to their own’ in Interpersonal relations, which was more difficult for elderly people; ‘He/she has access to new technologies’ in Personal development, which was more difficult for elderly people; ‘Technical aids are available if he/she needs them’; ‘He/she has healthy eating habits’ and ‘He/she maintains good personal hygiene’, which were more difficult for people at social disadvantage; ‘He/she has personal targets, goals and interests’ and ‘He/she chooses who they live with’ in Self-determination, the former is more difficult for the elderly whilst the latter is for people at social disadvantage; ‘He/she frequents communal areas’ in Social inclusion, was much more difficult for the elderly; and in Rights ‘One or more of his/her legal rights has been impaired’ was more difficult for people at social disadvantage. This meant that Physical wellbeing and Self-determination were the most problematic subscales in this sense. We thus note that 6 of the 10 items specified are more difficult for people at risk of social exclusion, whereby there could be a certain bias for this collective in these items. A more detailed discussion of these and other aspects is included in the final chapter of this work.

CHAPTER 6.

THE QUALITY OF LIFE OF USERS OF SOCIAL SERVICES IN CATALONIA

6. 1. Overall goal and hypotheses This chapter has been informed by the overall goal of the evaluation and analysis of the ‘objective’ quality of life of users of social services provided by ICASS in Catalonia. Accordingly, following the corroboration of the validity and reliability of the GENCAT Scale for such a purpose, we shall now list the main hypotheses underpinning the research process. Hypothesis 1. As in most studies on quality of life, the results in quality of life will be positive. Therefore, the scores in the overall scale and in each one of the dimensions will be higher than the scale’s theoretical mean point. Hypothesis 2. There are significant differences in the quality of life of the people evaluated according to several sociodemographic variables (gender, age, geographical location, marital status, place of residence and level of education). Hypothesis 3. The joint distribution of the scores recorded by the participants (divided into three groups: high, medium and low scores) and the sociodemographic variables do not fit the equiprobability model. Hypothesis 4. The scores recorded in the subscales allow predicting whether someone belongs to the groups of high, medium or low quality of life (according to the overall score recorded on the GENCAT Scale). 68

6. 2. Method

6. 2. 1. Participants The theoretical sample consisted of 3,029 users of 239 centres providing social services in Catalonia (attached to the Catalan Welfare and Social Services Institute - ICASS). The sample’s sociodemographic specifications are described in Chapter 4 herein.

6. 2. 2. Instrument The instrument used to evaluate the quality of life of the people involved in this research was the GENCAT Scale on quality of life (Verdugo et al., 2008a), a test whose design, validation and calibration have been described in preceding pages.

6. 2. 3. Procedure All the analyses used and described in this chapter have been conducted using the following statistics packages: SPSS 15.0 (SPSS, 2006), SAS 9.1.3 (SAS Institute, 2007), Statistica version 8.0 (StatSoft, 2007) and G*Power 3 (Faul, Erdfelder, Buchner & Lang, 2009) for Windows.

6. 3. Results Before proceeding with an analysis of the distributions and to the testing of the hypotheses considered above, we present a table (Table 7) with the descriptive statistics for each one of the ordinal variables evaluated. The distributions of both the scale in general and of each one of the dimensions were clearly asymmetric and negative, rejecting the hypotheses of univariate and multivariate normality in all cases. Furthermore, almost all the items recorded means of 3 and 4 in most cases, and both the means and the modes and medians always exceeded the scale’s theoretical mean points.

The participants recorded the highest scores in Rights and Material wellbeing, which is not surprising in our country (where civil and human rights are guaranteed) and in a sample such as the one in question (recipients of social services). In fact, within Material wellbeing the place of residence and the place of work are the aspects with the highest scores, whereas the lowest scores refer to income (although they are still positive, 33% provide a negative response). Elsewhere, the highest scores in Rights are recorded for the item ‘He/she is exposed to exploitation, violence or abuse’ (97% answer never or hardly ever), as well as for those items related to respect for rights by the organisation providing the social services. In this sense, there might be an element of bias in the answers as the questionnaires are filled in by professional staff working for these services. In turn, the areas with the greatest shortcomings in matters of Rights are related to the information users receive regarding their legal rights and their ability to defend them.

Table 7. Descriptive statistics of the scores on the GENCAT Scale Descriptors EW IR MW PD PW SD SI RI Total 23.479 27.077 28.887 21.498 27.800 24.116 23.046 35.868 211.693 Mean .092 .088 .052 .086 .052 .128 .071 .072 .466 S.E. of the Mean 2971 2875 2975 2982 2966 2968 2960 2955 2627 N valid 24 27 29 22 28 25 23 37 212 Median 25 28 32 23 29 27 23 40 209 Mode 238 252 645 262 431 166 323 652 49 Mode Frequency 8 10 14 8 17 9 9 17 136 Minimum 32 39 32 32 32 36 32 40 272 Maximum Theoretical Mean 16 20 16 16 16 18 16 20 138 Point 24.883 22.501 8.115 21.804 8.022 48.405 14.714 15.518 57.571 Variance 4.106 3.799 2.234 3.779 2.305 5.854 3.087 3.185 19.544 DT 24 29 18 24 15 27 23 23 136 Range 20 24 27 18 26 19 20 34 195 Percentiles 25 24 27 29 22 28 25 23 37 212 50 27 31 31 25 30 30 26 39 229 75 -.373 -.371 -1.067 -.154 -.549 -.213 -.137 -1.035 -.082 Asymmetry -.416 -.022 1.133 -.363 -.098 -.944 -.264 .713 -.453 Kurtosis NOTE. EW = Emotional wellbeing; IR = Interpersonal relations; MW = Material wellbeing; PD = Personal development; PW = Physical wellbeing; SD = Self-determination; SI = Social inclusion; RI = Rights.

We accept the first hypothesis, which posited that as in most studies the results obtained would be positive. A more unexpected result involves the very high scores recorded in Physical

wellbeing (the next highest after Rights), above all when we consider that amongst the participants there is a large group of disabled people and more than half are elderly (the majority are aged over 80). This suggests that the efforts made and work undertaken by the organisations charged with solving the problems most closely associated with care and improving health are successful, above all in the supervision of medication (96% answer always or almost always). The most problematic issues are those directly related to the user’s physical state: pain, discomfort or the inability to lead a normal life. The next ones are the dimensions of Interpersonal relations, Self-determination and Emotional wellbeing, in which the participants record average scores in comparison to other dimensions. In Interpersonal relations, we note that a very positive score is given to the user’s relationships with friends and family, whereas they are less so in relationships with partners and sexual relations. Regarding Self-determination, the item ‘The service he/she attends caters for their preferences’ records the highest scores, which leads us to suspect that there may once again be some form of bias in the answers. On the other hand, the aspects in which the users seem to show less self-determination are those involving the choice of whom they live with and taking decisions in their personal life, such as how to spend their money and what time they go to bed. Moreover, it is significant that almost half the sample (45%) do not organise their own lives. In Emotional wellbeing the most positive results are the low frequency of affirmative responses to those items related to feelings of insecurity, depression, anxiety and conduct issues, whereas somewhat lower responses were recorded for the items dealing with satisfaction (with oneself and with one’s present life), motivation at work and being happy and in a good mood. On the other hand, it seems that the people evaluated have the greatest problems (and therefore the lowest scores) in Personal development and Social inclusion. The highest score in Personal development is recorded in the work undertaken in the organisation providing social services (‘The service he/she attends caters for their personal development and the learning of new skills’). By contrast, the greatest difficulties involve access to technologies (only 16% do so always or almost always) and involvement in the design of their individual programme (only 18%). 71

Finally, one of the more positive results in Social inclusion is the low frequencies of affirmative responses to the item ‘He/she is rejected or discriminated against by others’ (4% answer always or almost always). In addition, high scores are recorded in encouraging participation in the various activities in the community by the service (54% reply always or almost always). The areas most lacking, however, are the use of communal areas, support from friends and friendships outside the service.

We accept the second hypothesis, according to which there are significant differences in quality of life depending on various sociodemographic variables. a. Gender: although we detect significant differences in favour of men in the overall

score (t(2582) = 2.57; p = .010), Emotional wellbeing (t(2582) = 3.61; p = .000), Physical wellbeing (t(2582) = 4.32; p = .000), Self-determination (t(2582) = 4.48; p = .000) and Personal development (t(2582) = 3.28; p = .001), these were very small and could largely be due to the high number of participants. b. Age: the analyses of variance gave rise to significant differences (p < .05) in all the scores except in Emotional wellbeing (p = .143). Nevertheless, post-hoc tests only confirmed these differences in the overall score, Self-determination, Social inclusion and Rights. Significantly lower scores than those for people aged under 20 were recorded by persons aged 31-40 in the overall score and those aged 61-70 in Social inclusion. However, in Self-determination and Rights those users aged between 71 and 80 recorded significantly higher scores than those aged 21-40. c. Geographical location: although the analysis of variance revealed significant differences between the various regions (Figure 7) in all the variables (p < .05), the corresponding post-hoc tests for uniform groups showed only two or more differentiated groups in terms of overall score (Tierras de Ebro, Ámbito de Poniente and Campo de Tarragona recorded significantly higher scores than Comarcas Gerundenses), Material wellbeing (Tierras de Ebro scores higher than Comarcas Gerundenses), Physical wellbeing (Comarcas Centrales records a significantly higher score than Alto Pirineo & Arán, Comarcas Gerundenses and Ámbito Metropolitano), Self-determination (Campo de Tarragona scores significantly higher than Alto Pirineo & Arán and Comarcas Gerundenses), Social inclusion (Tierras del Ebro scores significantly higher than Comarcas Gerundenses) and Rights (Campo de Tarragona scores significantly higher than Comarcas Gerundenses and Tierras del Ebro).

CATALONIA ALTO PIRINEO Y ARÁN

COMARCAS GERUNDENSES COMARCAS CENTRALES

ÁMBITO DE PONIENTE

ÁMBITO METROPOLITANO

CAMPO DE o TARRAGONA TIERRAS DEL EBRO

Figure 7. Regions of Catalonia

d. Condition: the condition variable gave rise to significant differences both in overall score (F(6) = 10.85; p = .000) and in all the dimensions (p = .000). This meant that people with intellectual disability record significantly lower scores overall (compared to people with substance abuse), as well as in Self-determination (compared to all the groups), in Personal development (compared to people with substance abuse, HIV or AIDS) and in Rights (compared to all the groups), whereas they record significantly higher scores in Emotional wellbeing (compared to people with HIV/AIDS), Material wellbeing (compared to people with Physical disability or HIV/AIDS) and Physical wellbeing (compared to all the other groups). This makes them the group with the highest fluctuations in the scores, with these differences suggesting shortcomings in the areas most directly related to abilities and (perhaps) a certain lack of realism or general satisfaction with their conditions of life. On the other hand, people with HIV/AIDS have the greatest difficulties in those aspects most closely associated with their social environment, as well as in satisfaction with their emotional and physical state. People with substance abuse, however, reveal the greatest problems in terms of Material wellbeing (which seems logical to us as they have recently had problems with substance abuse and are still in rehab). In addition, and as is to be expected given that they are not living in rest homes, elderly people in day centres record significantly higher scores in Interpersonal relations 73

(compared to people with HIV/AIDS and those with mental health problems) and Social inclusion (compared to people with HIV/AIDS and those with physical disability). Elderly people in homes tend to be at the more positive end with high scores except in Personal development (indeed, they record significantly lower scores than all the other groups apart from disabled people). This result is consistent with the situation of greater dependence of these people suffering from various physical and mental disorders associated with dementias and which, therefore, reduce their ability to learn new skills and also lead to the loss of skills already acquired. Elsewhere, people with mental health issues generally record high scores and are at the more positive end with significantly higher scores in Personal development, Selfdetermination, Social inclusion and Rights, whereas the more problematic areas appear to be Interpersonal relations. When we consider that many of these people have symptoms of depression, anxiety and other mental disorders, it comes as no surprise that this is the area with the lowest scores. Finally, people with physical disability tend to be at the end with the lowest scores, although they only record the most significantly low scores in Material wellbeing and Social inclusion, and the highest in Interpersonal relations. This suggests that there is still some work to be done with this collective regarding the existence of physical and social barriers. e. Marital status: this gave rise to significant differences in all the scores (p > .05). People who were separated recorded significantly lower scores in all the dimensions except for Self-determination and Rights (in which they recorded significantly higher scores). In Selfdetermination, single people without a partner, married people and widows/widowers recorded significantly lower scores, and in Rights single people are the ones in the worst conditions. By contrast, widows/widowers and single people without a partner record significantly higher scores in Material wellbeing (which is logical when we consider that they do not have to share expenses with a partner or pay alimony to an ex-partner). Single people without a partner record significantly higher scores than everyone else, whereby they are the ones with greater satisfaction both with themselves and with their lives. In turn, married people and single people with a partner and widows/widowers are the ones with the highest scores in those areas related to Interpersonal relations and Social inclusion, whereby it can be deduced that having a partner leads to a better social life. In Personal development, divorcees and single people with a partner record the highest scores, whilst single people without a partner do so in Physical wellbeing. Separated people and divorcees record significantly higher scores in Selfdetermination, which suggests they have greater abilities and opportunities for making choices

and taking decisions. Finally, widows/widowers, divorcees and separated people are the ones who best know, exercise and enjoy their legal and human rights. f. Place of residence: although the ANOVAS gave rise to significant differences in all the scores (p > .05), the post-hoc tests only confirmed the existence of several groups differentiated between each other in Physical wellbeing and Self-determination. This meant that people living in sheltered accommodation and independently without any support recorded significantly higher scores than the others in Physical wellbeing, and those who lived independently (both with and without support) recorded the highest scores in Selfdetermination. g. Level of education: as in the previous cases, the analyses of variance revealed significant differences in all the scores, but we only detected differentiated groups (in the posthoc tests) in Interpersonal relations, whereby the people who completed secondary education (‘Bachillerato’) or a basic degree (‘Diplomatura’) recorded significantly higher scores than those who studied the ‘Garantía Social’ programme of minimum basic education and vocational training.

We accept the third hypothesis that maintains that the joint distribution of the scores obtained (high, medium and low scores) and of the different sociodemographic variables does not fit the equiprobability model. As was to be expected, the results obtained when testing the data fit to the

independence model were consistent with those obtained when testing the preceding hypothesis. We thus noted the associations between the demographic variables and the scores summarised forthwith. a. Gender: we noted an association between being a man and recording high scores with a greater frequency than that expected in the overall scale, Self-determination, Personal development and Physical wellbeing. b. Age: there was a positive association between being aged over 80 and obtaining high scores in all the variables except in Physical wellbeing (in which it is those people aged 2150 who have that association). Those aged under 50 recorded low scores in Self-determination and Rights and those aged 61-70 in Physical wellbeing and Social inclusion. 75

c. Condition: belonging to the collective of elderly people was associated (as was to be expected) with lower-than-expected frequencies of high scores, although attendance of day centres is associated with positive scores in Social inclusion and Interpersonal relations, whereas living in rest homes is associated with good results in the overall score, Material wellbeing, Self-determination, Personal development and Rights. Having a physical disability is associated with favourable scores in Self-determination and unfavourable ones in Physical wellbeing, Social inclusion and Rights. By contrast, intellectual disability is associated with higher frequencies of unfavourable scores in the overall score, Interpersonal relations, Personal development, Self-determination and Rights, whereas these people record the highest scores in the three types of Wellbeing. Finally, problems of mental health, substance abuse and HIV/AIDS are associated with favourable scores in Personal development, Self-determination and Rights, whilst the first of these is associated with unfavourable scores in Emotional wellbeing, Physical wellbeing and Interpersonal relations; the second in Material wellbeing and the last in Wellbeing and Interpersonal relations. d. Marital status: having a partner was associated with positive scores in Interpersonal relations and negative ones in Material wellbeing. Being married, furthermore, was associated more frequently than expected with negative scores in Emotional wellbeing, Physical wellbeing and Personal development. Being widowed is associated more frequently than expected with unfavourable scores in Physical wellbeing (which is logical when we consider that most of the people are elderly) and being separated in Interpersonal relations and Material wellbeing (although being separated is associated more frequently than expected with positive scores in Personal development, Self-determination and Rights). Widowed people record higher frequencies in the high categories and lower frequencies in the low ones in several dimensions, whilst no significant association is made between this marital status and unfavourable scores in any case. Single people with no partner are the ones worst off as this status is most frequently associated with unfavourable scores. e. Place of residence: living in the family home had a significant association with less favourable scores in Emotional wellbeing, Physical wellbeing, Rights and Self-determination and with more favourable ones in Interpersonal relations and Social inclusion. Accordingly, living with one’s family seems to favour the more social areas. Living independently was associated with the highest scores in Interpersonal relations, Self-determination and Personal development (this last association was not observed for those who living independently with

support). Living in sheltered accommodation is associated with favourable scores in Social inclusion, Emotional wellbeing and Physical wellbeing and unfavourable ones in Personal development, Self-determination and Rights, a surprising situation when we consider that one of the main purposes of sheltered accommodation should precisely be the fostering of these dimensions. Finally, living in rest homes is associated with high scores in Self-determination and Rights and low ones regarding the more social areas (Interpersonal relations and Social inclusion), whereby as was to be expected living in a home seems to be associated with greater difficulties in one’s social life. f. Level of education: we found a significant association that was higher than expected between having studied primary and secondary education, a first or full degree and favourable scores, whereas having no formal schooling was associated with more unfavourable scores. Curiously, not having gone beyond secondary education had a significant and higherthan-expected association with less favourable results.

Finally, we rejected the fourth hypothesis, which posited that the scores obtained in the subscales would allow predicting whether the participants belonged to the groups of high, medium or low quality of life. The discriminant analyses made led us to conclude that the items in all the

dimensions best discriminate (i.e., highest percentage of cases correctly classified) those people with scores of medium quality of life, followed by people with low scores, with the worst results being for those with high scores. Nevertheless, the total percentage of cases correctly classified according to the score recorded in the items that make up each dimension was not overly low (between 53.10% and 73.05%), but it was not high enough for us to accept the hypothesis. Likewise, the multiple discriminant analyses made led us to the same conclusion for the eight dimensions. Nonetheless, the factor that returns the best results (i.e., best predictions or percentages of cases correctly classified) is Personal development (73.05%) and the one with the worst results is Material wellbeing (53.10% and 0% in the high category) and Physical wellbeing (55.80%). Nevertheless, we consider that rejecting this hypothesis is a positive result that provides further corroboration of the validity of the eight-dimension model given that if one of the dimensions were to correctly classify a very high percentage of people (close to 100%) there would be no need to evaluate all the other dimensions and it would mean that the model upon which the scale is based does not fit our data. 77

CHAPTER 7. DISCUSSION AND FUTURE LINES OF RESEARCH

7. 1. Specific conclusions The first main goal addressed in this study has involved developing a scale that will allow the objective evaluation of the quality of life of users of social services. There are two basic concepts underpinning this goal: quality of life and evaluation. Concerning the concept of quality of life, despite the fact that still today, after more than thirty years of research into the matter, we cannot affirm that there is an agreed definition, we find that the model proposed by Schalock & Verdugo (2002) is the one that appears to receive the greatest recognition in scientific literature both in Spain and abroad (e.g., Bonham et al., 2004; Gómez et al., 2008; Jenaro et al., 2005; Martín, 2006; Schalock, Bonham et al., 2008; Schalock, Verdugo et al., 2008; Schalock et al., 2005; van Loon et al., 2008; Verdugo, Arias & Gómez, 2006; Verdugo, Arias, Gómez & Schalock, 2008a, 2008b, in press; Verdugo, Gómez, Arias & Schalock, 2009; Verdugo, Gómez, Schalock & Arias, in press; Verdugo, Schalock et al., 2005, 2007). Although this model has emerged and been developed mainly within the field of disability, its application and acceptance are gradually spreading to other spheres and collectives (Alcedo et al., 2008; De Maeyer et al., 2009; Gómez et al., 2008; Verdugo et al., 2008b).

As regards evaluation, the assessment of the eight dimensions and of the central indicators that describe a person’s quality of life is undertaken through personal results that can be used to evaluate personal progress and guide strategies of organisational improvement (Keith & Bonham, 2005; Langberg & Smith, 2006; Schalock, Bonham & Verdugo, 2008; Schalock & Verdugo, 2007; Schalock, Verdugo et al., 2008; Veerman & van Yperen, 2007). Along these lines, there is general consensus regarding the types and uses of the evaluation of quality of life (Schalock, Bonham & Verdugo, 2008): depending on the purpose and perspective of the instrument developed, the indicators can be used for evaluating the wellbeing the person perceives (‘subjective’ evaluation) or the experiences and circumstances in a person’s life (‘objective’ evaluation). In the former case, data are gathered by self-reporting and it is the individual themselves who completes the test. In the latter case, in turn, the gathering of data is based on the direct observation of the individual’s conduct by a third person. In short, we can refer to objective and subjective evaluations and instruments according to their purpose, content and respondent. A review of the scientific literature has revealed that there are no instruments for evaluating the quality of life of users of social services and the efficacy of the interventions designed to improve them.

7. 1. 1. Building a quality of life scale 7. 1 .1. 1. Building a pool of items Given the instrument’s purpose (i.e., the evaluation of personal results and/or the efficacy of intervention programmes), we focused on the selection of items dealing with the microsystem and the mesosystem. The result at this stage was a bank of 106 items whose content alluded to one of the model's eight dimensions. The choice of the items was based on other Spanish and international instruments and research involving quality of life.

7. 1. 1. 2. Seeking expert opinion The following are the main conclusions to be drawn from the consultation of referees: 1) The mean agreement between experts is considered to range from good to excellent in terms of the assessment by both dimensions and by groups of referees. 2) The greatest consensus in the referee’s appraisals involved the appropriateness and importance of the items, whereas the agreement on observability was somewhat lower. 3) The referees’ mean rating in appropriateness, importance and observability was high (M = 7.852). 4) The closest agreements between the referees occur in the high scores in the three criteria. 5) The closest agreement between referees occurred in the dimension Material wellbeing, whereas there were greater discrepancies in the dimensions of Emotional wellbeing and Rights. 6) The experts in intellectual disability and mental health problems were the ones who rated the items with the highest scores, whilst the experts in sensorial disability were the ones awarding the lowest ratings. 7) The best results on the validity of the scale’s content were provided by the experts in intellectual disability. In our opinion, this circumstance is due to the fact that most of the items in the bank came from instruments targeting this collective and because the model upon which the scale is based has its origin and maximum development in this segment. This means that the experts in this collective are more familiar with the model and the type of items. 8) Both the agreement observed between referees and the high scores they awarded are validity evidence based on scale content.

7. 1. 1. 3. Discussion groups The main purpose behind the organisation of five discussion or focus groups was to meet the criteria related to engagement and consensus in the scale’s development amongst the main stakeholders in Catalonia. The involvement of professionals, users and family members associated with the social services targeted by the scale led to the corroboration of the results forthcoming from the consultation of the referees. The main conclusions were as follows: 1) Confirmation that the items were appropriate for the various collectives, a circumstance that helped to dispel doubts as to whether the content might be biased in favour of those people with intellectual disability. 2) The tweaking or redrafting of the items to enhance their understanding, define their content, or improve their applicability to the various collectives. 3) The proposal of new items that evaluated important aspects of the quality of life of the various collectives and which had not been considered in the initial pool of items. 4) Steps were taken to ensure the items adapted to the possible idiosyncrasies of the population attended to by the social services in Catalonia, so we can affirm that the scale has catered for and paid special attention to the ‘emic’ aspects of its target population. 5) The consensus reached by the members of each focal group regarding the appropriateness, importance and observability of the items, as well as by the various discussion groups, corroborated the validity of the instrument based on its content.

7. 1. 1. 4. Translation of the scale into Catalan The fact that from the very first the scale’s development was designed for a Catalan population and its adaptation was submitted for discussion in focus groups considerably helped this task, given that it was not a question of adapting it in the true sense of the word, as the scale’s content had already been tailored to suit the idiosyncrasies of the users of social services in Catalonia. We might therefore say that all that was required was a

simple translation of the items. Nevertheless, in order to uphold rigour in the process we followed all the guidelines recommended for the adaptation of tests by the International Tests Commission (ITC) (Bartram, 2001; Hambleton, 1993, 1994, 1996; Muñiz & Hambleton, 1996; van de Vyjver & Hambleton, 1996) and by the International Quality of Life Assessment Project (IQOLA) (Aaronson et al., 1992). The conclusion reached in that process is that the Catalan version of the scale is semantically, idiomatically and conceptually equivalent.

7. 1. 2. Validation of the GENCAT Scale by CTT and CFA

7. 1. 2. 1. Sample selection A stratified and polytypic sampling design was made according to the type of collective and the geographical location of the centre providing the services. In view of the characteristics of the population to be evaluated, one sampling was made for elderly people and another for people in situations of social disadvantage – which encompassed people with intellectual disability, physical disability, mental health problems, substance abuse and HIV/AIDS. It was estimated that there should be at least 1,110 people in each group, whereby the sampling errors were calculated to be 2.43% for the first group and 2.62% for the second one. Mention should be made at this point of the difficulties encountered in the sampling of people at social disadvantage, given that no updated census of the population was available. This meant it was not possible to calculate the sampling error for certain subgroups and that the total error was somewhat higher than for elderly people. This circumstance was compounded by other issues that did not involve the design of the research and which were more closely related to purely administrative or financial considerations: the budget catered solely for a representative sample in the case of the group of elderly people, whilst all the other collectives had to be gathered together in a diverse group that was called ‘people at social disadvantage’. In other words, we might venture to classify the sample of people at social disadvantage as representative of the population (with the noted proviso that there was no information available on the size of the population for certain collectives and, therefore, the sampling errors are a mere approximation); however, it would be totally naïve to conclude that each subgroup included here is representative of each one of the collectives in question. 82

7. 1. 2. 2. Participants and procedure The thorough procedure for collating data gave rise to an extensive sample made up of 3,029 people (1,619 elderly individuals and 1,410 individuals at social disadvantage) from 239 entities and evaluated by 608 professionals; a figure that far outstrips the one estimated in the sampling and means a considerable reduction in the sampling error. We should make special note here of the crucial role telephone contact played in ensuring a high participation in the study.

7. 1. 2. 3. Validity evidence based on content Concerning the results obtained regarding the validity evidence based on scale content, and in addition to those described beforehand from the consultation of experts and from the discussion groups, there are the following: 1)

All the tests made to verify whether the items discriminated between high, medium and low scores for quality of life were significant, although in general terms the scale seems to discriminate better between medium and low scores than amongst people with medium and high scores.

The dimension with the least discriminatory power is Rights. Physical wellbeing, Social inclusion, Interpersonal relations, Material wellbeing and Personal development have moderate discriminatory power. The subscales that most discriminate between people are Self-determination and Emotional wellbeing.

7. 1. 2. 4. Validity evidence based on internal structure In order to provide validity evidence based on internal structure, we conducted a CFA on the matrix of polychoric variances and covariances using the Diagonally Weighted Least Squares (DWLS) method for parameter estimation, given the absence of multivariate normality and the existence of atypical cases and extreme scores. Given the high number of items in the scale, we chose to conduct the CFA on parcels rather than on items.

Based on the theoretical review, we deemed it appropriate to compare the fit of our models to the different models forthcoming from recent research (Wang et al., in press) based on the multidimensional model by Schalock & Verdugo (2002). The results of the analysis of the fit to the various models led us to the following conclusions: 1) There is no suitable data fit to the model in which quality of life is understood to be a unidimensional concept. 2) There is no suitable data fit to the model by Wang. 3) There is a fit that could be considered acceptable to the Salamanca model. 4) There is a fit that could be considered suitable to the Schalock model. 5) The best data fit by far is to the multidimensional model by Schalock & Verdugo (2002). Accordingly, this was the theoretical model used to inform the structure of the GENCAT Scale in the eight dimensions or factors of quality of life. 6) If we make a careful study of the variance of the variables observed that explain the latent variables (R2), Self-determination is remarkable for the reliability of its indicators, followed by Emotional wellbeing and Personal development. The less reliable indicators in general terms, however, are included in Social inclusion. The remaining dimensions are characterised by combining indicators with low and moderate reliability.

7. 1. 2. 5. Reliability of the scale and subscales Reliability was assessed in terms of internal consistency and standard error of measurement both by dimensions and by collectives. The findings of these assessments enabled us to reach the following conclusions: 1) The internal consistency coefficients can be rated as high in the case of the overall scale. 2) The most reliable dimensions are Self-determination and Emotional wellbeing, whereas the least reliable dimension is Physical wellbeing. All the other dimensions record suitable coefficients.

3) There are no significant differences in the internal consistency coefficients (α) recorded for the various collectives in Physical wellbeing or Social inclusion, but there are in the six remaining dimensions, indicating that the items are not equally reliable in the different groups.

7. 1. 2. 6. Reliability of items With a view to assessing the reliability of the items, computation was made of their indices of difficulty and indices of reliability. A detailed study of these analyses led us to the following conclusions: 1) Generally speaking, the reliability of the items is adequate. The most reliable items are those in Self-determination and Emotional wellbeing, whereas the least reliable ones are in Physical wellbeing and Material wellbeing. The remaining dimensions combine highly reliable items with some of the least reliable ones. 2) The easiest items for the participants are the ones in Material wellbeing, Physical wellbeing and Rights, whilst the most difficult are the ones in Emotional wellbeing, Personal development and Self-determination. 3) There are a much higher number of items that are easy for the participants, as compared to the small number of items considered to be difficult.

7. 1. 3. Validation of the Scale using multigroup ESEM The exploratory structural equation analyses performed on the items confirmed the hypotheses regarding the fit of the eight-dimension model for the data on both elderly people and those at social disadvantage. However, the model’s fit was significantly better in the case of the elderly compared to those people at social disadvantage. The rejection of the hypothesis referring to the factor invariance between the groups suggests that the model does not behave in exactly the same way for the two groups.

One possible explanation for the model’s better fit with elderly people may lie in the greater uniformity of the data that we referred to earlier. We should remember that the definition ‘people at social disadvantage’ encompasses a wide range of collectives that, a priori, are extremely different to one another; for their part, however, elderly people constitute a much more uniform and numerous group that varied little in its responses.

7. 1. 4. Calibration of the GENCAT Scale with the RSM It is worth stressing what, in general terms, we believe to be the main contribution of the analyses made from the perspective of the TRI: not only do they corroborate the results obtained through CTT, but they also act as a complement by shedding some light on their possible interpretations or explanations. There follows an overview of some of the analyses' more specific conclusions: 1) In general terms, the data fit the model. The items with a more questionable fit are ‘He/she has a satisfactory sex life’ in Interpersonal relations and ‘He/she is exposed to exploitation, violence or abuse’. 2) The highest point-biserial correlations (whose interpretation is similar to the alpha coefficient in the CTT) were recorded in Self-determination and Emotional wellbeing. The lowest appeared in Physical wellbeing. 3) The dimensions Rights and Material wellbeing record the most extreme scores (i.e., very high scores). 4) Most of the items are considered to be very accurate: 59 (85.5%) have a measurement error of .02 to .03. 5) The most accurate items (i.e., whose observed score is closest to the person’s true score in the construct evaluated) are in the dimensions Self-determination, Emotional wellbeing, Personal development and Social inclusion (in descending order). On the other hand, the least accurate ones are (in ascending order) in Rights, Physical wellbeing, Material wellbeing and Interpersonal relations.

6) There is a major discrepancy between the excellent reliability or separation of the items and the moderate or low reliability of the individuals. This means it would be convenient to include a number of more difficult items in these dimensions in order to raise their level of accuracy (not just simply to increase their number, but rather to adjust the items’ difficulty to the level of competence). 7) Ratification is largely made of the result obtained through the CTT regarding the items’ ease or difficulty: easy items prevail over difficult ones. 8) The calibration of the items shows that the dimensions Interpersonal relations, Selfdetermination and Rights cover a greater range of personal results compared to all the others. This phenomenon could indeed be explained by the greater number of items included in them. 9) The dimension Interpersonal relations is the one with the greatest suitability in terms of the difficulty of the items for the sample. 10) The response categories are generally suitable. Nevertheless, the extreme categories (‘never or hardly ever’ and ‘always or almost always’) are the ones used most compared to the intermediate ones. The option ‘often’ is used much more than ‘sometimes’ (i.e., for the items of positive valence; and the opposite occurs for those of negative valence). The latter, furthermore, does not seem to be appropriate for Interpersonal relations, Material wellbeing and Rights, given that there are few opportunities for it to be used in these dimensions. 11) As a whole, the items and dimensions provide the most information in intermediate ability levels, whilst the level of information drops in the lower levels. 12) There is no differential item functioning (DIF) by gender with the exception of i39 (He/she maintains good personal hygiene) in Physical wellbeing which, a priori and strictly applying the assumptions of the RSM, should be deleted due to a possible bias in favour of women; that is, it appears to measure men and women differently, whereby it might reflect a content more closely associated with social aspects of men and women instead of with attributes of Physical wellbeing.

13) There is a differential function in 10 of the 69 items (14.5%) by condition or collective, with the sole exception of Material wellbeing, where there do not appear to be any biases. We thus find that 6 of the 10 items specified are more difficult for the people at risk of social exclusion and the other four are easier for them than for elderly people. This points to a possible bias in favour of one collective or the other. Nevertheless, a qualitative analysis of the items with differential function does not permit us to single out any explanation or characteristic that might be detrimental to one of the groups and which is not directly related to those people’s quality of life, so we conclude that they are not biased items.

7. 1. 5. The quality of life of the users of social services in Catalonia 1) The results of the evaluation of quality of life involving the users of social services in Catalonia are positive, as is the case in most studies. 2) There are significant differences in the quality of life of the users according to several sociodemographic variables (gender, age, collective, marital status, place of residence and level of education). 3) The scores recorded in the subscales allow predicting whether someone belongs to the groups of high, medium or low quality of life according to the overall score obtained in the GENCAT Scale. Nevertheless, as the TRI already indicated, the items in all the dimensions discriminate better those people with average scores, followed by those with low scores. The scale discriminates worse amongst people with high scores, which might be explained by the simplicity (i.e. ease) of the items in some of the dimensions. The dimensions that discriminate least are Material wellbeing and Physical wellbeing, whereas on this occasion Personal development is the one that does so the best.

7. 2. General conclusions

As the main conclusion of this work, we should note that this study is a first and unprecedented approach to the evaluation of the quality of life of the users of social services. Rather than simply concluding that it is a valid and reliable instrument, it is our understanding that we have provided validity evidence for the GENCAT Scale that is more than satisfactory and based on: (a) its content and adaptation to the Catalan population (experts, discussion groups, discriminatory power); (b) its factor structure (CFA, multigroup ESEM; and (c) people's individual responses (RSM). In terms of reliability, we consider an acceptable level has been reached by most of the dimensions; however, Physical wellbeing has a highly questionable reliability. Accordingly, it would be convenient to review the content of this subscale, with there probably being a need to discard some of the listed items (or move them to other more appropriate dimensions) and include others more closely linked to the construct. We have also noted that an improvement could be made to the accuracy and reliability of subscales such as Material wellbeing and Rights, including more difficult items for this population. In turn, the dimensions Self-determination and Emotional wellbeing can be considered the most appropriate ones. A further highlight is the fact that both the GENCAT Scale and the quality of life model underpinning it appear to function better and provide better results with the elderly instead of with the collective of people at social disadvantage. This finding leads us to consider the need to increase the number of participants in the various subgroups included here in order to confirm this differential performance or, in the best of cases, ratify that it is due simply to this group's major diversity. Finally, although we should view the results obtained in Physical wellbeing with some caution, we can conclude that the quality of life of social services users is fairly high. The results are excellent in Material wellbeing and Rights. By contrast, it seems advisable to dedicate more effort to improving Personal development and Social inclusion.

7. 2. 1. The study’s strong points We should like to draw special attention to the rigorous process followed in the development of the GENCAT Scale, as well as to the efforts made by the research team in its formulation. Accordingly, we now present what we consider to be the strong points of this research. Independently of the results obtained and although it is a first approach to the evaluation of the quality of life of users of social services, we can affirm that little research has hitherto been conducted in this field. Furthermore, we should like to stress the combination of methods particular to CTT and TRI (and within these, the use of methodologies that have been developed only very recently, such as ESEM analyses), which imbue this work with a high degree of innovation. Firstly, note should be taken of the extensive number of expert referees (uncommon in research such as that conducted here) and the subsequent organisation of discussion groups. We understand that this has provided more than sufficient validity evidence based on scale content. Secondly, in order to ensure the sample was representative and the results could be generally applicable to the population, a demanding process was undertaken to provide a sample that was stratified by collectives and geographical areas. The difficulty this entailed was due not only to the nature of the task itself but also to the many gaps and errors in the censuses available in Catalonia, which meant that we were forced to undertake the added task of updating them. A third aspect worthy of mention was the decision to resort not only to e-mail but also to telephone contact to seek the involvement of as many centres and individuals as possible. Finally, we were exceptionally fortunate to receive the cooperation of most of the centres selected, and we believe that telephone contact played a crucial role in this. Fourthly, we can affirm that the GENCAT Scale has sufficient evidence regarding its validity for evaluating the quality of life of users of social services in Catalonia and it stands as the only instrument so far available that is sensitive to those intervention programmes designed to improve personal results (and therefore the quality of life of users) and so can be of considerable use when applied in the services and in the development and assessment of programmes. Indeed, the results forthcoming comfortably

outperform those provided by other instruments designed to evaluate quality of life. Nonetheless, we do not dismiss other kinds of complementary evaluation; quite the contrary, we consider it highly recommendable to apply it together with, for example, subjective evaluations of quality of life (e.g., INTEGRAL Scale; Verdugo, Gómez et al., 2009), supports (Supports Intensity Scale, SIS; Verdugo, Arias & Ibáñez, 2007), adaptive behaviour (Diagnostic Adaptive Behavior Scale, DABS; Verdugo, Arias & Navas, 2009), and Self-determination (Verdugo, Vicent & Gómez, 2006), etc. 7. 2. 2. The study’s limitations Although the work has several strong points, it is not without its limitations. In addition to those already singled out throughout the text itself and those mentioned earlier in the conclusions and which we shall not cite to avoid repetition, we can note those weak points detailed forthwith. Firstly, in spite of the efforts made in the selection of the sample and during the data gathering process, we can only claim that the sample is representative in the case of elderly people. The lack of information on the sizes of the population in all the other subgroups, as well as grouping them together according to the common denominator of people at social disadvantage, considerably diminished the representative nature of this part of the sample. Although the reasons for grouping such diverse collectives together as one were out of our hands – due to budgetary considerations – this is nonetheless one of the study’s major limitations. For this reason, we deem it necessary to confirm the scale’s suitability with representative samples of the various collectives, which would require increasing the size of this part of the sample. Secondly, an important limitation was the fact that we had no data on the reliability between evaluators, an issue that is of vital importance for a scale in which quality of life is not evaluated by the actual individual involved but by third parties and, further still, when we rate it as objective. Although there are now studies – as yet unpublished – along these lines that are promising, their validation is one of the priority goals from here on. Likewise, it would be convenient to prove the reliability of the test-retest for verifying the temporal reliability of the scores.

Thirdly, the absence of other instruments for evaluating the quality of life of users of social services, as well as having no knowledge of an external criterion on quality of life, rendered it impossible to provide validity evidence referred to the criterion. In this sense, an extensive line of future research seems to involve an analysis of the correlation between the scores in the GENCAT Scale and those recorded in other tests designed for specific populations (e.g., the INTEGRAL Scale for disabled people and the FUMAT Scale for the elderly). Along these lines, van Loon et al. (2008) conducted a study on the psychometric properties of the Personal Outcomes Scale using the GENCAT Scale as a criterion, which means that some evidence is available in this sense (Claes et al., 2009). Finally, despite the thorough and rigorous development process and the fact that generally speaking the results are satisfactory, another of the instrument’s limitations is the low reliability of the subscale Physical wellbeing and the excessive ease of the items in Material wellbeing and Rights, which mean that ensuing studies should consider the inclusion of more difficult items in the latter two subscales and a review of the items in the first one.

7. 2. 3. Future lines of research Regarding more specific future lines of research that are more directly related to the GENCAT Scale, and in addition to those already mentioned for overcoming this study’s limitations, we should like to single out four additional lines that complement the preceding ones. We are currently focusing our efforts on pursuing these goals.

Firstly, we consider it extremely important to verify the instrument’s reliability and validity for the Spanish population. To this end, its application has already begun in several autonomous communities, or regions, to see if they mirror the excellent results recorded with users of social services in Catalonia. Secondly, given its usefulness and its satisfactory psychometric properties, it would be of great interest to adapt the scale for application in other countries. We currently have a version translated into English and another one translated into Dutch. The latter is being used as the criterion in quality of life evaluations performed in the Netherlands with the Personal Outcomes Scale (van Loon et al., 2008).

Thirdly, the study of the relationship between the objective and subjective evaluation of quality of life is still today a field of study with contradictory results that we consider to be of special significance and which merits further research (e.g., via the GENCAT Scale and the INTEGRAL Scale by Verdugo, Gómez et al., 2009). Finally, the line of research on quality of life opens up a wide range of potential areas for investigation when we consider how relevant this construct can be today to others of such importance (especially within the field of intellectual disability), such as the concepts of supports (e.g., via the Supports Intensity Scale, SIS; adapted by Verdugo, Arias & Ibáñez, 2007) and adaptive behaviour (e.g., via the Diagnostic Adaptive Behavior Scale, DABS; adapted by Verdugo, Arias & Navas, 2009). At this point, we should remember that this work is framed within a line of research on quality of life that was begun more than a decade ago by the University of Salamanca’s Institute on Community Integration (INICO), headed by its director, Dr Miguel Ángel Verdugo, when he began to cooperate with Dr Robert L. Schalock on the conceptualisation of the construct and the best ways of evaluating it. Over the course of these years there have been numerous and varied publications (e.g., Arias et al., in press; Schalock et al., in press; Verdugo, Arias, Gómez & Schalock, in press; Verdugo, Gómez, Schalock & Arias, in press), research projects (e.g., BSO2003-03059 and SEJ2006-12575 of the Spanish Ministry of Science and Technology; GR193, Excellence Group of the regional government, the Junta, of Castilla y León, amongst others), doctoral theses (e.g., Córdoba, 2004; Crespo, 2003; Gómez-Vela, 2003; González, 2002; Ibáñez, 2009; Martín, 2006; Morentin, 2008; Sabeh, 2004), degree dissertations (e.g., Gómez, 2005), training courses, congresses and scientific symposia on evaluation instruments (Gómez et al., 2008; Verdugo & Schalock, 2001; Verdugo, Arias, Gómez & Schalock, 2008a; Verdugo, Gómez & Arias, in press; Verdugo, Gómez, Arias & Schalock, 2009), etc., which have been induced by this line of research. Since the identification and study of the main indicators of the quality of life dimensions of the multidimensional model proposed by Schalock in 1996 (which we referred to in the theoretical background presented in this work), there have been transcultural studies designed to verify the factor structure and corroborate the construct’s etic and emic features (Jenaro et al., 2005; Schalock et al., 2005) and the model’s suitability has been confirmed in several studies and using a broad range of methodologies (e.g., confirmatory factor analysis, structural equations and neural networks).

Having an increasingly sound and internationally accepted model and suitable instruments for its valid and reliable evaluation has meant that today we are now one step on from the conceptualisation and evaluation of quality of life. We can say that we are now at just the right moment to undertake its evaluation in the applied field and in several spheres. Accordingly, we now see the main line of research to be its application in social services with a view to helping to develop and evaluate programmes designed to improve personal results and quality of life. We therefore consider the GENCAT Scale to be a significant contribution in this sense. We can use it to address what up until now has been the foremost difficulty arising in our field, namely, the real application of the concept in social services. Thus, we now have an instrument sensitive to changes that will enable us to make a valid evaluation of personal results as a criterion for identifying needs and designing programmes, at the same time as we also monitor progress in the inclusive process and in the planning of supports focusing on the individual (Schalock, Gardner & Bradley, 2007; Verdugo, 2009). We therefore look upon the GENCAT Scale as a unique and extraordinary instrument that is extremely useful for driving change in professional practices and in the personal results obtained by people at greater risk of social exclusion (microsystem), in organisations (mesosystem) and in social policies (macrosystem). We should like to end by expressing our vehement hope that this study, despite its limitations and admitting that there is still a lot of work to be done, will at least be one little step forward towards the achievement of the changes mentioned at the level of microsystem, mesosystem and, why not, macrosystem.

REFERENCES Aaronson, N.K., Acquadro, C., Alonso, J., Apolone, G., Bucquet, D., Bullinger, M., Bungay, K., Fukuhara, S., Gandek, B., Keller, S. et al. (1992). International Quality of Life Assessment (IQOLA) Project [PDF]. Quality of Life Research, 1(5), 349-51. Recuperado de http://www.jstor.org/pss/4035108. Abdel-Kader, K., Unruh, M. L., & Weisbord, S. D. (2009). Symptom burden, depression, and quality of life in chronic and end-stage kidney disease. Clinical Journal of the American Society of Nephrology, 4, 1057-1064. doi: 10.2215/CJN.00430109 Aguado, A. L., Alcedo, M. A., Rueda, M. B., Flórez, M. A., & Gutiérrez, M. (1994). Una escala multidimensional de evaluación de personas con lesión medular: primeros resultados en Asturias y Euskadi. Paper presented at IV Congreso de Evaluación Psicológica, september, Santiago de Compostela, Spain. Aguado, A. L., Alcedo, M. A., Rueda, M. B., Santos, A. M., & González, I. (1997, march). Una aproximación psicométrica a la Escala Multidimensional de Evaluación de Lesionados Medulares. Paper presented at II Jornadas Científicas de Investigación sobre

Personas

con

Discapacidad,

Salamanca,

Spain.

Retrieved

from

http://campus.usal.es/~inico/investigacion/jornadas/jornada2/simpos/s14.htm. Aguado, A. L., González, I., Alcedo, M. A., & Arias, B. (2003). Calidad de vida y lesión medular. In M. A. Verdugo & B. Jordán de Urríes (Coords.), Investigación, innovación y cambio (pp. 624-664). Salamanca: Amarú. Alcedo, M. A., Aguado, A. L., Arias, B., González, M., & Rozada, C. (2008). Escala de Calidad de Vida (ECV) para personas con discapacidad que envejecen: estudio preliminar.

Intervención

Psicosocial,

17,

153-167.

Retrieved

from

http://scielo.isciii.es/pdf/inter/v17n2/v17n2a05.pdf Anderson, K. L., & Burckhardt, C. S. (1999). Conceptualization and measurement of quality of life as an outcome variable for health care intervention and research. Journal of advanced nursing, 29, 298-306. doi: 10.1046/j.1365-2648.1999.00889.x Andrich, D. (1978). Application of a psychometric rating model to ordered categories which are scored with successive integers. Applied Psychological Measurement, 2, 581-594. doi: 10.1177/014662167800200413 Angermeyer, M. C., Holzinger, A., Matschinger, H., & Stengler-Wenzke, K. (2002). Depression and quality of life: results of a follow-up study. The International Journal of Social Psychiatry, 48, 189-99. doi: 10.1177/002076402128783235 Arias, B. (2008). Desarrollo de un ejemplo de análisis factorial confirmatorio con LISREL, AMOS y SAS. In M. A. Verdugo, M. Crespo, M. Badía & B. Arias (Coords.). Metodología en la investigación sobre discapacidad. Introducción al uso de las ecuaciones estructurales. Salamanca, Spain: INICO.

107

Arias, B., Gómez, L. E., Verdugo, M. A., & Navas, P. (in press). Evaluación de la calidad de vida en drogodependientes mediante el modelo de Rasch. Revista Española de Drogodependencias. Arostegui, I. (2002). Calidad de vida y retraso mental. Bilbao, Spain: Mensajero. Asparouhov, T., & Muthén, B. (in press). Exploratory structural equation modeling. Structural

Equation

Modeling.

Retrieved

from

http://www.statmodel.com/

download/EFACFA810.pdf Aznar, A. S., & Castañón, G. G. (2005). Quality of life from the point of view of Latin American families: a participative research study. Journal of Intellectual Disability Research, 49, 784-788. Bartram, D. (2001). Guidelines for test users: a review of national and international initiatives. European Journal of Psychological Assessment, 17, 173-186. Batista-Foguet, J. M., & Coenders, G. (2000). Modelos de ecuaciones estructurales. Madrid: Hespérides. Batista-Foguet, J. M., Coenders, G., & Alonso, J. (2004). Análisis Factorial Confirmatorio. Su utilidad en la validación de cuestionarios relacionados con la salud. Medicina Clínica,

122,

21-27.

Retrieved

from

http://www3.udg.edu/fcee/professors/

gcoenders/pap21.pdf Bollen, K. A. (1989). Structural equations with latent variables. New York: Wiley. Bonham, G. S., Basehart, S., & Marchand, C. B. (2003). Ask me! FY2003 report: the quality of life of Marylanders with developmental disabilities receiving DDA funded support. Baltimore,

Maryland:

The

ARC

Maryland.

Retrieved

from

http://

www.bonhamresearch.com/PDF/2003TBasehart%20Ask%20Me %20FY2003Sum.pdf Bonham, G. S., Basehart, S., & Marchand, C. B. (2006). Ask Me! FY 2006: The QOL of Marylanders with developmental disabilities receiving DDA funded supports. Annapolis, Maryland: Bonham Research. Bonham, G. S., Basehart, S., Schalock, R. L., Marchand, C. G., Kirchner, N., & Rumenap, J. M. (2004). Consumer-based quality of life assessment: The Maryland Ask Me! Project. Mental Retardation, 42, 338-355. doi: 10.1352/0047-6765(2004)42<338: CQOLAT>2.0.CO;2 Borthwick-Duffy, S. A. (1992). Quality of life and quality of care in mental retardation. In L. Rowitz (Ed.), Mental retardation in the year 2000 (pp. 52-66). Berlin: SpringerVerlag. Brown, I., Keith, K. D., & Schalock, R. L. (2004). Quality of life conceptualization, measurement, and application: Validation of the SIRG-QOL consumer principles. Journal of Intellectual Disability Research, 48, 451-451.

108

Brown, R. I., Schalock, R. L., & Brown, I. (2009). Special issue: quality of life and its applications. Journal of Policy and Practice in Intellectual Disability, 6, 1-1. doi: 10.1111/j.1741-1130.2009.00204.x Browne, M. W. (1982). Covariance structures. In D.M. Hawkins (Ed.), Topics in applied multivariate analysis (pp. 72-141). Cambridge, UK: Cambridge University Press. Browne, M. W. (1984). Asymptotically distribution-free methods in the analysis of covariance structures. British Journal of Mathematical and Statistical Psychology, 37, 62-8. Browne, M. W. (2001). An overview of analytic rotation in exploratory factor analysis. Multivariate Behavioral Research, 36, 111–150. Retrieved from http://www.psych. umn.edu/faculty/waller/classes/mult09/readings/brownembr.pdf Byrne, B. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS. Mahwah, New Jersey: Earlbaum. Caballo, C., Crespo, M., Jenaro, C., Verdugo, M. A., & Martínez, J. L. (2005). Factor structure of the Schalock and Keith Quality of Life Questionnaire (QOL-Q): validation on Mexican and Spanish samples. Journal of Intellectual Disability Research, 49, 773-776. doi: 10.1111/j.1365-2788.2005.00750.x Carretero-Dios, H., & Pérez, C. (2007). Standards for the development and review of instrumental studies: considerations about test selection in psychological research. International Journal of Clinical and Health Psychology, 7, 863-882. Retrieved from http://www.aepc.es/ijchp/carretero_es.pdf Carta, A., Braccio, L., Belpoliti, M., Soliani, L., Sartore, F., Gandolfi, S. A., & Maraini, G. (1998). Self-assessment of the quality of vision: association of questionnaire score with

objective

clinical

tests.

Current

Eye

Research,

17,

506-511.

doi:

10.1076/ceyr.17.5.506.5191 Cea, M. A. (2002). Análisis multivariable. Teoría y práctica en la investigación social. Madrid: Síntesis. Chou, Y. C., & Schalock, R. L. (2009). Survey outcomes and cross-national comparisons of quality of life with respect to people with intellectual disabilities in Taiwan. Journal of Policy and Practice in Intellectual Disabilities, 6, 7-10. doi: 10.1111/j.17411130.2008.00200.x Claes, C., van Hove, G., van Loon, J., Vandevelde, S., & Schalock, R. L. (2009). Quality of Life Measurement in the Field of Intellectual Disabilities: Eight Principles for Assessing Quality of Life-Related Personal Outcomes. Social Indicator Research. doi: 10.1007/s11205-009-9517-7 Córdoba, L. (2004). Satisfacción con la calidad de vida en familias de niños y adolescentes con discapacidad (Unpublished doctoral dissertation). University of Salamanca, Salamanca, Spain.

107

Cragg, R., & Harrison, J. (1985). Living in a supervised home. A questionnaire of quality of life. Manchester, UK: West Midlands Campaign for People with Mental Handicap. Crespo, M. (2003). Validación y aplicación de la escala de calidad de vida de Schalock y Keith (1993) a adultos con discapacidad intelectual en contextos laborales (Unpublished doctoral dissertation). University of Salamanca, Salamanca, Spain. Cummins, R. A. (1993). The Comprehensive Quality of Life Scale for Intellectual Disability (4th ed.) Melbourne, Australia: Deakin University. Cummins, R. A. (1996). Quality of life and disability. In P.O´Brien & R. Murray (Eds.), Human services: towards partnership and support (pp. 255-268). Auckland, NZ: Dunmore. Cummins, R. A. (1997). Assessing quality of life. In I. Brown (Ed.), Quality of life for people with disabilities (2ª ed.) (pp. 116-150). Cheltenham, UK: Stanley Thornes. Cummins, R. A. (2000). Objective and subjective quality of life: An interactive model. Social Indicators Research, 52, 55-72. doi: 10.1023/A:1007027822521 Cummins, R. A. (2005). Moving from the quality of life concept to a theory. Journal of Intellectual Disability Research, 49, 699-706. doi: 10.1111/j.1365-2788.2005.00738.x Cunningham, W. E., Crystal, S., Bozzette, S., & Hays, R. D. (2005). The association of healthrelated quality of life with survival among persons with HIV infection in the United States. Journal of general internal medicine, 20(1), 21-7. doi: 10.1111/j.15251497.2005.30402.x. De Maeyer, J., Vanderplasschen, W., & Broekaert, E. (2009). Exploratory study on drug users’ perspectives on quality of life: more than health-related quality of life? Social Indicators Research, 90, 107-126. doi: 10.1007/s11205-008-9315-7 EuroQOL Group (1990). EuroQOL – A new facility for the measurement of health related quality of life. Health Policy, 16, 199-208. Farquhar, M. (1995). Definitions of quality of life: A taxonomy. Journal of Advanced Nursing, 22, 502–508. Faul, F., Erdfelder, E., Buchner, A., & Lang, A. G. (2009). G*Power 3.0.10 [programa informático].

Retrieved

from

http://www.psycho.uniduesseldorf.de/abteilungen /aap/gpower3/gpower31. Felce, D., & Perry, J. (1995). Quality of life: Its definition and measurement. Research in Developmental Disabilities, 16, 51-74. doi: 10.1016/0891-4222(94)00028-8 Felce, D., & Perry, J. (1996). Assessment of quality of life. In R. L. Schalock (Ed.), Quality of life, vol. I: Conceptualization and measurement (pp. 63-73). Washington, DC: American Association on Mental Retardation. Felce, D., & Perry, J. (1997). Quality of life: The scope of the term and its breadth of measurement. In R. I. Brown (Ed.), Quality of life for persons with disabilities: Models, research and practice (2nd ed.) (pp. 56-71). Cheltenham, UK: Stanley Thornes.

108

Fischer, B., Rehm, J., & Kim, G. (2001a). Quality of Life (QoL) in illicit drug addiction treatment and research: concepts, evidence and questions. In B. Westermann, C. Jellinek & G. Belleman (Eds.), Substitution: Zwischen Leben und Sterben (pp. 21–40). Weilheim, Germany: Beltz Deutscher Studien Verlag. Fischer, B., Rehm, J., & Kim, G. (2001b). Whose quality of life is it, really? British Medical Journal, 322, 1357–1360. Frost, N. A., Sparrow, J. M., Durant, J. S., Donovan, J. L., Peters, T. J., & Brookes, S. T. (2008). Development of a questionnaire for measurement of vision-related quality of life. Ophthalmic Epidemiology, 5, 185-210. doi: 10.1076/opep.5.4.185.4191 García, L. (2006). La perspectiva de género en el proceso de adaptación a la lesión medular (doctoral

dissertation).

Retrieved

from

http://www.tesisenxarxa.net/

TESIS_UOV/AVAILABLE/TDR-0423107-101846/UOV0002s.pdf?ref=starshemale. com Gardner, J. F., & Carran, D. (2005). Attainment of personal outcomes by people with developmental disabilities. Mental Retardation, 43, 157-174. doi: 10.1352/00476765(2005)43[157:AOPOBP]2.0.CO;2 Giner, J., Ibáñez, E., Franco, D., & Alarcón, R. D. (2008). Subjectivity in the perception and measurement of quality of life: conceptualization and development of the Seville questionnaire. Psyquiatric Quarterly, 79, 321-30. doi: 10.1007/s11126-008-9086-y Gómez, L. E. (2005). Construcción de una escala de calidad de vida para adultos con discapacidad intelectual (Unpublished dissertation). University of Salamanca, Salamanca, Spain. Gómez, L. E., Verdugo, M. A., Arias, B., & Navas, P. (2008). Evaluación de la calidad de vida en personas mayores y con discapacidad: la Escala FUMAT. Intervención psicosocial, 17, 189-200. Gómez-Vela, M. (2003). Evaluación de la Calidad de Vida en Alumnos de ESO con NEE y sin ellas (Unpublished doctoral dissertation). University of Salamanca, Salamanca, Spain. Gómez-Vela, M., & Sabeh, E. (2000). Calidad de vida. Evolución del concepto y su influencia en la investigación y en la práctica. Integra, 9, 1-4. Retrieved from http://www3.usal.es/~inico/investigacion/invesinico/calidad.htm. González, F. (2002). Calidad de vida percibida por los niños hospitalizados de Castilla y León (Unpublished doctoral dissertation). University of Salamanca, Salamanca, Spain. Goodley, D., Armstrong, D., Sutherland, K., & Laurie, L. (2003). Self-advocacy, learning difficulties, and the social model of disability. Mental Retardation, 43, 149-160. doi: 10.1352/0047-6765(2003)41<149:SLDATS>2.0.CO;2 Gothwal, V. K., Wright, T., Lamourex, E. L., & Pesudovs, K. (2009). Rasch Analysis of the Quality of Life and Vision Function Questionnaire. Optometry and Vision Science, 86, 836-844.

107

Hambleton, R. K. (1993). Translating achievement tests for use in cross-national studies. European Journal of Psychological Assessment, 9, 57-68. Retrieved from http://www. eric.ed.gov/ERICDocs/data/ericdocs2sql/content_storage_01/0000019b/80/13/d5/41 .pdf Hambleton, R. K. (1994). Guidelines for adapting educational and psychological tests: a progress report. European Journal of Psychological Assessment, 10, 229-244. Hambleton, R. K. (1996). Adaptación de tests para su uso en diferentes idiomas y culturas: fuentes de error, posibles soluciones y directrices prácticas. In J. Muñiz (Coor.), Psicometría. Madrid: Universitas. Harner, C. J., & Heal, L. W. (1993). The Multifaceted lifestyle satisfaction scale: Psychometric properties for an interview schedule for assessing personal satisfaction for adults with limited intelligence. Research in Developmental Disabilites, 14, 221-236. doi: 10.1016/0891-4222(93)90032-F Heal, L. W., & Chadsey-Rusch, J. (1985). The Lifestyle Satisfaction Scale (LSS): assessing individuals' satisfaction with residence, community setting, and associated services. Applied Research in Mental Retardation, 6, 475-90. Heckman, T. G. (2003). The chronic illness quality of life (CIQOL) model: explaining life satisfaction in people living with HIV disease. Health Psychology, 22, 140-147. Herrman, H. & Chopra, P. (2008). Quality of life and neurotic disorders in general healthcare. Current Opinion in Psychiatry, 22, 61-68. doi: 10.1097/YCO. 0b013e32831a4750 Hewitt, J. (2007). Critical evaluation of the use of research tools in evaluating quality of life for people with schizophrenia. International Journal of Mental Health Nursing, 16, 214. doi: 10.1111/j.1447-0349.2006.00438.x Hughes, C., & Hwang, B. (1996). Attempts to Conceptualize and Measure Quality of Life. In R. L. Schalock (Ed.), Quality of Life. Vol. I: Conceptualization and Measurement (pp. 51–61). Washington, DC: AAMR. Hughes, C., Hwang, B., Kim, J., Eisenman, L. T., & Killian, D. J. (1995). Quality of life in applied research: A review and analysis of empirical measures. American Journal on Mental Retardation, 99, 623-641. Hunt, S. M., McEwen, J., & McKenna, S. P. (1985). Measuring health status: a new tool for clinicians and epidemiologists. The Journal of the Royal College of General Practitioners, 35, 185-188. Retrieved from http://www.pubmedcentral.nih.gov/ picrender.fcgi?artid=1960139&blobtype=pdf Ibáñez, A. (2009). Adaptación de la Escala de Intensidad de Apoyos (SIS): adecuación psicométrica y estructura factorial (Unpublished doctoral dissertation). University of Salamanca, Salamanca, Spain.

108

Janssen, C. G. C., Schuengel, C., & Stolk, J. (2005). Perspectives on quality of life of people with intellectual disabilities: the interpretation of discrepancies between clients and caregivers. Quality of Life Research, 14, 57-69. doi: 10.1007/s11136-004-1692-z Jenaro, C., Verdugo, M. A., Caballo, C., Balboni, G., Lachapelle, Y., Otbrebski, W., & Schalock, R. L. (2005). Cross-cultural study of person-centered quality of life domains and indicators: a replication. Journal of Intellectual Disability Research, 49, 734-739. doi: 10.1111/j.1365-2788.2005.00742.x Kaplan,

(2000).

Structural

equation

modeling:

foundations

and

extensions.

Psychometrika, 54, 681-697. Keith, K. D. (2007). Quality of life. In A. Carr, G. O’Reilly, P. N. Walsh & J. M. Cevoy (Eds.), The handbook of intellectual disability and clinical psychology practice (pp. 143-168). London, UK: Routledge. Keith, K. D., & Bonham, G. S. (2005). The use of quality of life data at the organization and systems level. Journal of Intellectual Disability Research, 49, 799-805. doi: 10.1111/j.1365-2788.2005.00755.x Keith, K. D., & Ferdinand, L. R. (2000). Project to compare quality of life of Nebraskans with developmental disabilities and citizens without disabilities. Lincoln, NB: Governor’s Planning Council on Developmental Disabilities. Keith, K. D., Heal, L.W., & Schalock, R. L. (1996). Cross-cultural measurement of critical quality of life concepts. Journal of Intellectual and Developmental Disability, 21, 273293. doi: 10.1080/13668259600033201 Keith, K. D., & Schalock, R. L. (2000). Cross-cultural perspectives on quality of life. Washington, DC: American Association on Mental Retardation. Kline, R. B. (2005). Principles and practice of structural equation modeling. New York: The Guilford Press. Kober, R., & Eggleton I. R. C. (2002). Factor stability of the Schalock and Keith (1993) Quality of Life Questionnaire. Mental Retardation, 40, 157–65. doi: 10.1352/00476765(2002)040<0157:FSOTSA>2.0.CO;2 Koch, L., Rumrill, P., Roessler, R., & Fitzgerald, S. (2001). Illness and demographic correlates of quality of life among people with multiple sclerosis. Rehabilitation Psychology, 46, 154-164. La Grow, S. J. (2007). Predicting Perceived Quality of Life Scores from the National Eye Institue 25-Item Visual Function Questionnaire. Optometry & Vision Science, 88, 785-788. doi: 10.1097/OPX.0b013e31812f5f24 Langberg, J. & Smith, B. (2006). Developing evidence-based interventions for deployment into school settings: A case example highlighting key issues of efficacy and effectiveness. Evaluation and Program Planning, 29, 323-334.

107

Levasseur, M., Desrosiers, J., & Tribble, D. (2008). Do quality of life, participation and environment of older adults differ according to level of activity? Health Quality of Life Outcomes; 6, 30-30. Retrieved from http://www.pubmedcentral.nih.gov/ articlerender.fcgi?artid=2412860 Linacre, J. M. (2002). What do infit and outfit, mean square and standardized mean? Rasch Measurement Transactions, 12(2), 878. Retrieved from http://www.rasch.org/rmt/ rmt162f. html. Linacre, J. M. (2008). Winsteps. Rasch measurement computer program (Computer software). Chicago, IL: Winsteps.com. Linacre, J. M. & Wright, B. D. (1999). A user's guide to WINSTEPS, BIGSTEPS, MINISTEP Rasch model computer programs. Chicago; IL: MESA Press. Loehlin, J. C. (2004). Latent variable models: An introduction to factor, path, and structural equation analysis. Mahwah, NJ: Lawrence Erlbaum Associates. Mangione, C. M., Lee, P. P., Gutierrez, P. R., Spritzer, K., Berry, S., & Hays, R. D. (2001). Development

the

25-item

National

Eye

Institute

Visual

Function

Questionnaire. Archives of Ophthalmology, 119, 1050-1058. Retrieved from http://archopht.ama-assn.org/cgi/content/full/119/7/1050?ck=nck Marcoulides, G. A., & Schumacker, R.E. (Eds.) (2001). New developments and techniques in structural equation modeling. Mahwah, NJ: Lawrence Erlbaum Associates. Marsh, H. W., & Hocevar, D. (1994). Confirmatory factor analysis models of factorial invariance: A multifaceted approach. Structural Equation Modeling 1(10), 5-34. Martín, J. C. (2006). Modelo de calidad de vida de Schalock y Verdugo (2002-2003): Validación mediante análisis multivariable (Unpublished doctoral dissertation). University of Salamanca, Salamanca, Spain. Messick, S. M. (1975). The standard problem. Meaning and values in measurement and evaluation. American Psychologist, 29, 955-966. Messick, S. M. (1980). Test validity and the ethics of assessment. American Psychologist, 35, 1012-1027. doi: 10.3102/0013189X018002005 Messick, S. M. (1988). The once and future issues of validity: assessing the meaning and consequences of measurement. In H. Wainer & H., Braun (Eds.), Test validity. Hillsdale, NJ: Lawrence Erlbaum Associates. Messick, S. M. (1989). Validity. In R. Linn (Ed.), Educational measurement (pp. 13-103). New York: Macmillan. Messick, S. M. (1995). Standards of validity and the validity of standards in performance assessment.

Educational

Measurement:

10.1111/j.1745-3992.1995.tb00881.x

108

Issues

and

Practice,

14,

5-8.

doi:

Messick, S. M. (1996). Validity of performance assessment. In G. W. Phillips (Ed.), Technical issues in large-scale performance assessment (pp. 1-19). Washington, DC: National Center for Educational Statistics. Messick, S. M. (1998). Test validity a matter of consequence. Social Indicator Research, 45, 3544. doi: 10.1023/A:1006964925094 Moore, M., Höfer, S., McGee, H., & Ring, L. (2005). Can the concepts of depression and quality of life be integrated using a time perspective? Health Quality of Life Outcomes, 3, 1-1. doi: 10.1186/1477-7525-3-1 Morentin, R. (2008). Amor y calidad de vida en personas con discapacidad intelectual (Unpublished doctoral dissertation). University of Salamanca, Salamanca, Spain. Muñiz, J., & Hambleton, R. K. (1996). Directrices para la traducción y adaptación de los tests.

Papeles

del

Psicólogo

66,

63-70.

Retrieved

from

http://www.papelesdelpsicologo.es/ vernumero.asp?id=737 Muthén, L. K., & Muthén, B. O. (2008). Mplus statistical software (Computer software). Los Angeles, CA: Muthén & Muthén. Navarro, D., García-Heras, S., Carrasco, O., & Casas, A. (2008). Calidad de vida, apoyo social y deterioro en una muestra de personas con trastorno mental grave. Intervención psicosocial, 17, 321-336. Nota, L., Soresi, S., & Perry, J (2006). Quality of life in adults with an intellectual disability: the Evaluation of Quality of Life Instrument. Journal of Intellectual Disability Research, 50, 371-385. doi: 10.1111/j.1365-2788.2006.00785.x Ouellette-Kuntz, H., McCreary, B. D., Minnes, P. M., & Stanton, B. (1994). Evaluating Quality of Life: The Development of the Quality of Life Interview Schedule (QUOLIS). Journal on Developmental Disability, 3,17-31. Perry, J., & Felce, D. (1995). Objective assessments of quality-of-life—How much do they agree with each other. Journal of Community & Applied Social Psychology, 5, 1–19. doi: 10.1002/casp.2450050102 Perry, J., & Felce, D. (2005). Factors associated with outcomes in community group homes. American Journal on Mental Retardation, 110, 121-135. doi: 10.1352/08958017(2005)110<121:FAWOIC>2.0.CO;2 Petry, K., Maes, B., & Vlaskamp, C. (2005). Domains of quality of life of people with profound multiple disabilities: the perspective of parents and direct support staff. Journal of Applied Research in Intellectual Disabilities, 18, 35-46. doi: 10.1111/j.14683148.2004.00209.x Petry, K., Maes, B., & Vlaskamp, C. (2007). Support characteristics associated with the quality of life of people with profound intellectual and multiple disabilities: the perspective of parents and direct support staff. Journal of Policy and Practice in Intellectual Disabilities, 4, 104–110. doi: 10.1111/j.1741-1130.2007.00107.x

107

Popham, W. J. (2000). Modern educational measurement. Practical guidelines for educational leaders. Boston, MA: Allyn & Bacon. Prieto, G., & Muñiz, J. (2000). Un modelo para evaluar la calidad de los tests utilizados en España.

Papeles

del

Psicólogo,

77,

65-75.

Retrieved

from

http://www.papelesdelpsicologo.es/vernumero.asp?id=1102 Pukrop, R., Möller, H. J., & Steinmeyer, E. M. (2000). Quality of life in psychiatry: a systematic contribution to construct validation and the development of the integrative assessment tool "modular system for quality of life". European Archives of Psychiatry & Clinical Neuroscience, 250, 120-32. Rapley, M., & Lobley, J. (1995). Factor analysis of the Schalock and Keith (1993) Quality of Life Questionnaire: a replication. Mental Handicap Research, 8, 194-202. doi: 10.1111/j.1365-2788.2005.00750.x Raykov, T., & Marcoulides, G.A. (2008). An introduction to applied multivariate analysis. New York: Taylor & Francis. Renwick, R., & Brown, I. (1996). The centre for health promotion’s conceptual approach to quality of life: Being, belonging, and becoming. In R. Renwick & I. Brown (Eds.), Quality of life in health promotion and rehabilitation: Conceptual approaches, issues, and applications (pp. 75–86). Thousand Oaks, CA: Sage. Renwick, R., Nourhaghighi, N., Manns, P., & Rudman, D. B. (2003). Quality of life for people with physical disabilities: a new instrument. International Journal of Rehabilitation Research, 26, 279-288. Roussel, P., Durieu, F., Campoy, E., & El Akremi, A. (2002), Méthodes d'équations structurelles: Recherche et applications en gestion. Paris: Economica. Rudman, D., Renwick, R., Raphael, D., & Brown, I. (1995). The quality of life profile for adults with physical disabilities. Canadian Journal of Occupational Therapy, 62, 2525. Rueda, B. (2001). Evaluación del afrontamiento en lesionados medulares. (Unpublished doctoral dissertation). University of Deusto, Bilbao, Spain. Rueda, B., Aguado, A. L., & Alcedo, M. A. (2008). Estrés, afrontamiento y variables psicológicas intervinientes en el proceso de adaptación a la Lesión Medular (LM): una revisión de la bibliografía. Intervención psicosocial, 17, 109-124. Retrieved from http://scielo.isciii.es/pdf/inter/v17n2/v17n2a02.pdf Ryu, E., West, S. G., & Sousa, H. (2009). Mediation and Moderation: Testing Relationships Between Symptom Status, Functional Health, and Quality of Life in HIV Patients. Multivariate Behavioral Research, 44, 213-232. doi: 10.1080/00273170902794222 Sabeh, E. (2004). Evaluación de la calidad de vida en la infancia: niños con necesidades educativas especiales y niños de escolaridad común. Un estudio comparativo (Unpublished doctoral dissertation). University of Salamanca, Salamanca, Spain.

108

SAS Institute (2007). SAS v.9.13. (Computer software). Cary, NC: SAS Institute Inc. Satorra, A., & Bentler, P. M. (1994). Corrections to test statistics and standard errors in covariance structure analysis. In A.Von Eye & C.C. Clogg (Eds.), Latent variables analysis: Applications for developmental research (pp. 399-419). Thousand Oaks, CA: Sage. Schalock, R. L., & Bonham, G. S. (2003). Measuring outcomes and managing for results. Evaluation & Program Planning, 26, 229-235. doi: 10.1016/S0149-7189(03)00027-2 Schalock, R. L., Bonham, G. S., & Verdugo, M. A. (2008). The concept of quality of life as a framework for program planning, implementation, evaluation, and improvement. Evaluation and Program Planning, 3, 181-190. doi:10.1016/j.evalprogplan.2008. 02.001 Schalock, R. L., Brown, I., Brown, R., Cummins, R. A., Felce, D., Matikka, L., Keith, K. D., & Parmenter, T. (2002). Conceptualization, measurement, and application of quality of life for persons with intellectual disabilities: Results of an international panel of experts. Mental Retardation, 40, 457-470. doi: 10.1352/0047-6765(2002)040<0457 :CMAAOQ>2.0.CO;2 Schalock, R. L., Gardner, J. F., & Bradley, V. J. (2007). Quality of life for people with intellectual and other developmental disabilities. Applications across individuals, organizations, communities, and systems. Washington, DC: American Association on Intellectual and Developmental Disabilities. Schalock, R. L., Keith, K. D., Verdugo, M. A., & Gomez, L. E. (in press). Quality of life model development in the field of intellectual disability. In R. Kober (Ed.), Quality of life for people with intellectual disability. New York: Springer. Schalock, R. L., & Felce, D. (2004). Quality of life and subjective well-being: conceptual and measurement issues. In E. Emerson, C. Hatton, T. Thompson & T. R. Parmenter (Eds.), International handbook of applied research in intellectual disabilities (pp. 261279). London: John Wiley & Sons. Schalock, R. L., & Keith, K. D. (1993). Quality of Life Questionnaire. Worthington, OH: IDS Publishing. Schalock, R. L., & Verdugo, M. A. (2002). Quality of life for human service practitioners. Washington, DC: American Association on Mental Retardation. Schalock, R. L., & Verdugo, M. A. (2007). El concepto de calidad de vida en los servicios y apoyos para personas con discapacidad intelectual. Siglo Cero, 38, 21-36. Schalock, R. L., & Verdugo, M. A. (2008). Quality of life: from concept to application in the field of intellectual disabilities. Evaluation & Program Planning, 31, 181-190. Schalock, R. L., Verdugo, M. A., Bonham, G. S., Fantova, F., & Van Loon, J. (2008). Enhancing personal outcomes: organizational strategies, guidelines, and

107

examples. Journal of Policy and Practice in Intellectual Disabilities, 5, 18-28. doi: 10.1111/j.1741-1130.2007.00135.x Schalock, R. L., Verdugo, M. A., Jenaro, C., Wang, M., Wehmeyer, M., Xu, J., & Lachapelle, Y. (2005). A cross-cultural study of quality of life indicators. American Journal on Mental Retardation, 110, 298-311. doi: 10.1352/0895-8017(2005)110[298:CSOQOL] 2.0.CO;2 Scientific Software International (2006). LISREL, v.8.8 (Computer software). Lincolnwood, IL: Scientific Software International. Shogren, K. A., Bradley, V. J., Gomez, S. C., Yeager, M. H. et al. (en prensa). Public policy and the enhancement of desired public policy outcomes for persons with intellectual disability. Intellectual and Developmental Disabilities. Skevington, S. (2002). Advancing cross-cultural research on quality of life: observations drawn from the WHOQOL development. Quality of Life research, 11, 135-144. doi: 10.1023/A:1015013312456 SPSS (2006). SPSS, V. 14.0 (Computer software). Chicago, IL: SPSS Inc. Statsoft, Inc. (2007). Statistica 8.0 (Computer software). www.statsoft.com. Stewart, A. L., & Ware, J. E. (1992). Measuring Functioning and Well-Being: The Medical Outcomes Study Approach. Durham, NC: Duke University Press. Tamarit, J. (2005). Hacia un Sistema de Evaluación de la Calidad en FEAPS. Intervención psicosocial, 14, 295-308. Van de Vijver, F., & Hambleton, R. K. (1996). Translating tests: Some practical guidelines. European Psychologist, 1, 89-99. Van Loon, J. (2008). Aligning quality of life domains and indicators to support intensity scale data. In R. L. Schalock, J. R. Thompson & M. J. Tasse (Eds.), Supports intensity scale companion guide: A resource for SIS users (pp. 80-87). Washington, DC: American Association on Intellectual and Developmental Disabilities. Van Loon, J. (2009). Uso de la Escala de Intensidad de Apoyos en la práctica. Asignación de recursos y planificación de apoyos individuales basados en la Escala de Intensidad de Apoyos. Siglo Cero, 40, 54-66. Van Loon, J., van Hove, G., Schalock, R. L., & Claes, C. (2008). Personal Outcomes Scale. Middleburg, Holland: Arduin Steichlich. Veenhoven, R. (1996). Happy life-expectancy a comprehensive measure of quality-of-life in nations. Social Indicators Research, 39, 1-58. Retrieved from http://www2.eur.nl/ fsw/research/veenhoven/Pub1990s/96b-full.pdf Verdugo, M. A. (2006). Cómo mejorar la calidad de vida de las personas con discapacidad. Salamanca: Amarú. Verdugo, M. A. (2009). Calidad de vida, I+D+i y políticas sociales. Siglo Cero, 40, 5-21.

108

Verdugo, M. A., Arias, B., & Gómez, L. E. (2006). Escala integral de medición subjetiva y objetiva de la calidad de vida en personas con discapacidad intelectual. In M.A. Verdugo (Dir.), Como mejorar la calidad de vida de las personas con discapacidad. Instrumentos y estrategias de evaluación (pp. 417-448). Salamanca, Spain: Amarú. Verdugo, M. A., Arias, B., Gómez L. E., & Schalock, R. L. (2008a). Formulari de l’Escala Gencat de Qualitat de vida. Manual d’aplicació de l’Escala Gencat de Qualitat de vida. Barcelona: Departamento de Acción Social y Ciudadanía, Generalitat de Cataluña. Verdugo, M. A., Arias, B., Gómez L. E., & Schalock, R. L. (2008b). Escala GENCAT. Informe sobre la creació d’una escala multidimensional per avaluar la qualitat de vida de les persones usuàries dels serveis socials a Catalunya. Barcelona: Departamento de Acción Social y Ciudadanía, Generalitat de Cataluña. Verdugo, M. A., Arias, B., Gómez L. E., & Schalock, R. L. (2009). Formulario de la Escala GENCAT de Calidad de vida. Manual de aplicación de la Escala GENCAT de Calidad de vida. Barcelona: Departamento de Acción Social y Ciudadanía, Generalitat de Cataluña. Verdugo, M. A., Arias, B., Gómez L. E., & Schalock, R. L. (in press). Development of an objective instrument to assess quality of life in social services: Reliability and validity in Spain. International Journal of Clinical and Health Psychology. Verdugo, M. A., Arias, B., Gomez, L. E. & van Loon, J. (2007). Schaal Voor Kwaliteit Van Bestaan Voor Mensen Met Een Verstandelijke Beperking. Holland: Stichting Arduin. Verdugo, M. A., Arias, B., & Ibáñez, A. (2006). La Escala de Intensidad de Apoyos: Un instrumento para evaluar y planificar las necesidades de apoyo de adultos con discapacidad intelectual. In M. A. Verdugo (Dir.), Cómo mejorar la calidad de vida de las personas con discapacidad. Instrumentos y estrategias de evaluación (pp. 475-498). Salamanca, Spain: Amarú. Verdugo, M. A., Arias, B., & Navas, P. (2009). Adaptación y validación al castellano de la Escala de Diagnóstico de Conducta Adaptativa (DABS) de la AAIDD: Estudio piloto. In M. A. Verdugo, T. Nieto, B. Jordán de Urríes & M. Crespo (Eds.), Mejorando resultados personales para una vida de calidad (pp. 531-540). Salamanca, Spain: Amarú. Verdugo, M. A., Gómez, L. E., & Arias, B. (2007). La Escala INTEGRAL de calidad de vida: desarrollo y estudio preliminar de sus propiedades psicométricas. Siglo Cero 38, 37-56. Verdugo, M. A., Gómez, L. E., & Arias, B. (in press). La Escala FUMAT de evaluación de la calidad de vida de personas mayores y con discapacidad. Salamanca, Spain: INICO. Verdugo, M. A., Gómez, L. E., Arias, B., & Schalock, R. L. (2009). Escala Integral de Calidad de Vida. Madrid. CEPE.

107

Verdugo, M. A., Gómez, L. E., Schalock, R. L., & Arias, B. (in press). The Integral Quality of Life Scale: Development, Validation, and Use. In R. Kober (Ed.), Quality of life for people with intellectual disability. New York: Springer. Verdugo, M. A., & Martín, M. (2002). Autodeterminación y calidad de vida en salud mental: dos conceptos emergentes. Salud Mental, 25, 68-87. Retrieved from http://redalyc. uaemex.mx/redalyc/src/inicio/ArtPdfRed.jsp?iCve=58242507 Verdugo, M. A., Prieto, G., Caballo, C., & Peláez, A. (2005). Factorial Structure of the Quality of Life Questionnaire in a Spanish sample of visually disabled adults. European Journal of Psychological Assessment, 21, 44-55. doi: 10.1111/j.13652788.2005.00750.x Verdugo, M. A., & Schalock, R. L. (2001). El concepto de calidad de vida en los servicios humanos. In M. A. Verdugo & F. B. Jordán de Urríes (Coords.), Apoyos, autodeterminación y calidad de vida (pp. 105-112). Salamanca, Spain: Amarú. Verdugo, M. A., Schalock, R. L., Gómez, L. E., & Arias, B. (2007). Construcción de escalas de calidad de vida multidimensionales centradas en el contexto: la Escala GENCAT. Siglo Cero, 38, 57-72. Verdugo, M. A., Schalock, R. L., Keith, K., & Stancliffe, R. (2005). Quality of life and its measurement: Important principles and guidelines. Journal of Intellectual Disability Research, 49, 707-717. doi: 10.1111/j.1365-2788.2005.00739.x Verdugo, M. A., Vicent, C., & Gómez, L. E. (2006). Un instrumento para medir la autodeterminación en adultos con discapacidad. In M. A. Verdugo (Dir.), Cómo mejorar la calidad de vida de las personas con discapacidad. Instrumentos y estrategias de evaluación (pp. 449-474). Salamanca: Amarú. Walsh, P., Erickson, E., Bradley, V., Moseley, C., & Schalock, R. L. (2006). Supported accommodation services for people with intellectual disabilities: A review of models and instruments used to measure quality of life in various settings. Dublin: National Disability Authority. Wang, M., Schalock, R. L., & Verdugo, M. A. (in press). The Cross-Cultural Validation and Analysis of Quality of Life Construct Validity. Manuscript submitted for publication. Ware, J. E. J., & Sherbourne, C. D. (1992). The MOS 36-item short-form health survey (SF36), I: conceptual framework and item selection. Medical Care, 30, 473-483. doi: 10.1177/016327879702000102 Wehmeyer, M., & Kelchner, K. (1995). The Arc's self-determination scale: Adolescent version. Arlington, TX: The ARC. Wisniewski, S. R., Rush, A. J., Bryan, C., Shelton, R., Trivedi, M. H., Marcus, S., Husain, M. M., Hollon, S. D., & Fava, M. (2007). Comparison of Quality of Life Measures in a Depressed Population. The Journal of Nervous and Mental Disease, 195, 219-225. Retrieved from http://redalyc.uaemex.mx/redalyc/src/inicio/ArtPdfRed.jsp?iCve= 58242507

108

Wolfensberger, W., & Glenn, L. (1975). Program analysis of service systems: Handbook and Manual (3rd ed.) Toronto: National Institute on Mental Retardation. Wolfensberger, W., & Thomas, S (1983). PASSING (Program Analysis of Service Systems' Implementation of Normalization Goals): Normalization criteria and ratings manual (2nd ed.). Toronto: National Institute on Mental Retardation. Wolffsohn, J. S., & Cochrane, A. L. (2000). Design of the low vision quality of life questionnaire (LVQOL) and measuring the outcome of low-vision rehabilitation. American Journal of Ophthalmology, 130, 793-802. doi: 10.1016/S0002-9394(00)006103 Wright, B. D., & Masters, G. N. (1982). Rating scale analysis. Chicago, IL: MESA Press. Xu, J., Wang, M., Xiang, Y., & Hu, X. (2005). Quality of life for people with intellectual disabilities in China: a cross-culture perspectives study. Journal of Intellectual Disabilities Research, 49, 745-749. doi: 10.1111/j.1365-2788.2005.00744.x Yates, A. (1987). Multivariate exploratory data analysis: A perspective on exploratory factor analysis. New York: State University of New York Press. Zumbo, B. D. (2007). Validity: Foundational issues and statistical methodology. In C. R. Rao & S. Sinharay (Eds.), Handbook of statistics, Vol. 26: Psychometrics (pp. 45–79). Amsterdam, Holland: Elsevier Science B.V.

107

APENDIX 1. ENGLISH VERSION OF THE GENCAT SCALE

EMOTIONAL WELLBEING 1.

He/she is satisfied with their present life

He/she shows symptoms of depression

He/she is happy and in a good mood

He/she expresses feelings of helplessness or insecurity

He/she shows symptoms of anxiety

He/she is satisfied with themselves

He/she has problems of conduct

He/she is motivated when performing some kind of activity

INTERPERSONAL RELATIONS 9.

He/she does things they enjoy with other people

10. The relations with his/her family are as they would like them to be 11. He/she complains about a lack of close friends 12. He/she has a negative view of their friendships 13. He/she says they feel undervalued by their family 14. He/she finds it difficult to start up a relationship with a potential partner 15. He/she gets on well with their colleagues at work 16. He/she says they feel loved by the people who are important to them 17. Most of the people with whom they interact are in a similar situation to their own 18. He/she has a satisfactory sex life

107

MATERIAL WELLBEING 19. Where he/she lives stops them from leading a healthy life (noise, fumes, odours, gloom, lack of ventilation, damage, inaccessibility…) 20. His/her workplace complies with rules on health and safety 21. He/she has the material possessions they need 22. He/she is unhappy with where they live 23. Where he/she lives is clean 24. He/she has enough money to cover their basic needs 25. He/she does not earn enough to be able to afford luxuries 26. Where he/she lives has been adapted to their needs

PERSONAL DEVELOPMENT 27. He/she finds it difficult to cope with everyday situations 28. He/she has access to new technologies (Internet, mobile phone, etc.) 29. The work they do enables them to learn new skills 30. He/she finds it difficult to effectively deal with the problems they have to face 31. He/she does their work competently and responsibly 32. The service he/she attends caters for their personal development and the learning of new skills 33. He/she is involved in the drafting of their own individual programme 34. He/she lacks motivation at work

PHYSICAL WELLBEING 35. He/she finds it difficult to sleep 36. Technical aids are available if he/she needs them 37. He/she has healthy eating habits 38. His/her state of health allows them to lead a normal life 39. He/she maintains good personal hygiene 40. The service he/she attends supervises the medication they take 41. His/her health problems cause them pain and discomfort 42. He/she finds it difficult to access healthcare resources (preventive care, GP, at home, in hospital, etc.)

108

SELF-DETERMINATION 43. He/she has personal targets, goals and interests 44. He/she decides how to spend their free time 45. The service he/she attends caters for their preferences 46. He/she defends their ideas and opinions 47. Other people decide upon his/her personal life 48. Other people decide how he/she spends their money 49. Other people decide what time he/she goes to bed 50. He/she organises their own life 51. He/she chooses who they live with

SOCIAL INCLUSION 52. He/she frequents communal areas (public swimming pools, cinemas, theatres, museums, libraries…) 53. His/her family provides support whenever needed 54. There are physical, cultural or social barriers that hinder his/her social inclusion 55. He/she lacks the necessary support for taking an active part in everyday life in their community 56. His/her friends provide support whenever it is needed 57. The service he/she attends encourages them to take part in community activities 58. The only friends he/she has are the ones who attend the same service 59. He/she is rejected or discriminated against by others

107

RIGHTS 60. His/her family violates their privacy (reading their letters, entering without knocking…) 61. He/she is treated with respect in their environment 62. He/she has information on their basic rights as a citizen 63. He/she finds it difficult to defend their rights when these are violated 64. The service he/she attends respects their privacy 65. The service he/she attends respects their possessions and their ownership rights 66. One or more of his/her legal rights have been impaired (citizenship, vote, legal processes, respect for their beliefs, values, etc.) 67. The service he/she attends respects and defends their rights (confidentiality, information on their rights as a user…) 68. The service respects the privacy of his/her information 69. He/she is exposed to exploitation, violence or abuse

108