STEP 1:
• State the hypotheses
Null hypothesis • denoted by H0 • a statement that the value of a population parameter is equal to some claimed value • the prediction that there is no interaction between variables
Alternative hypothesis • denoted by H1 or Ha or HA • a statement that the parameter has a value that somehow differs from the null hypothesis • the prediction that there is a measurable interaction between variables
Effect the bio-fertilizer ‘x’ on plant growth
Differences Example đ?‘Żđ?&#x;Ž • đ?‘Żđ?&#x;Ž : đ?? = đ?? đ?&#x;Ž • đ?‘Żđ?&#x;Ž : đ?? ≼ đ?? đ?&#x;Ž • đ?‘Żđ?&#x;Ž : đ?? ≤ đ?? đ?&#x;Ž
Example đ?&#x2018;Żđ?&#x2018;¨ â&#x20AC;˘ đ?&#x2018;Żđ?&#x;&#x17D; : đ?? â&#x2030; đ?? đ?&#x;&#x17D; â&#x20AC;˘ đ?&#x2018;Żđ?&#x;? : đ?? < đ?? đ?&#x;&#x17D; â&#x20AC;˘ đ?&#x2018;Żđ?&#x;? : đ?? > đ?? đ?&#x;&#x17D;
Null Hypothesis: đ??ť0
â&#x20AC;˘ Must contain condition of equality: =, ď&#x201A;ł, or ď&#x201A;Ł â&#x20AC;˘ We wish to reject â&#x20AC;˘ Assume true until proven otherwise.
Alternative Hypothesis: đ??ťđ??´
â&#x20AC;˘ The symbolic form of the alternative hypothesis must use one of these symbols: ď&#x201A;š, <, >. â&#x20AC;˘ Must be true if H0 is false â&#x20AC;˘ â&#x20AC;&#x2DC;oppositeâ&#x20AC;&#x2122; of Null â&#x20AC;˘ We support â&#x20AC;˘ We trying to prove by conducting the inferential statistics
Type of alternative hypothesis
One-tailed (directional) â&#x20AC;˘ Right-tail â&#x20AC;˘ Left-tail
Two-tailed (non-directional)
Differences Directional
Onetailed
• predicts the actual DIRECTION • more precise • relies on previous studies
Twotailed
• predicts an OPEN outcome • very general • no other research has been done
• One-tailed
Non-directional • Two-tailed
Examples One-tailed hypothesis
• Example 1
• The national mean cholesterol level is appproximately 210. • 100 people with high cholesterol levels (over 265) participated in a drug study and were treated with a new drug Cholestyramine. • After treatment the sample mean was 228 and the sample standard deviation was 12. • One question of interest is whether people taking this drug still have a mean cholesterol level that exceeds the national average. • What are the null and alternative hypotheses?
Examples One-tailed hypothesis â&#x20AC;˘ đ?&#x2018;Żđ?&#x;&#x17D; : đ?? = đ?? đ?&#x;&#x17D; â&#x20AC;˘ đ?&#x2018;Żđ?&#x2018;¨ : đ?? > đ?? đ?&#x;&#x17D;
â&#x20AC;˘ Solution đ?&#x2018;Żđ?&#x;&#x17D; : đ?? = đ?&#x;?đ?&#x;?đ?&#x;&#x17D; đ?&#x2018;Żđ?&#x2018;¨ : đ?? > đ?&#x;?đ?&#x;?đ?&#x;&#x17D;
â&#x20AC;˘ Population Characteristic: â&#x20AC;˘ Îź = Average cholesterol level for all people taking this drug
Examples Two-tailed hypothesis
• Example 2 • We have a medicine that is being manufactured and each pill is supposed to have 14 milligrams of the active ingredient. • What are our null and alternative hypotheses?
Examples Two-tailed hypothesis â&#x20AC;˘ đ?&#x2018;Żđ?&#x;&#x17D; : đ?? = đ?? đ?&#x;&#x17D; â&#x20AC;˘ đ?&#x2018;Żđ?&#x2018;¨ : đ?? â&#x2030; đ?? đ?&#x;&#x17D;
â&#x20AC;˘ Solution đ?&#x2018;Żđ?&#x;&#x17D; : đ?? = đ?&#x;?đ?&#x;&#x2019; đ?&#x2018;Żđ?&#x2018;¨ : đ?? â&#x2030; đ?&#x;?đ?&#x;&#x2019; â&#x20AC;˘ Our null hypothesis states that the population has a mean equal to 14 milligrams. â&#x20AC;˘ Our alternative hypothesis states that the population has a mean that is different than 14 milligrams.
Possible outcomes Reject đ?&#x2018;Żđ?&#x;&#x17D; and accept đ?&#x2018;Żđ?&#x2018;¨ â&#x20AC;˘ sufficient evidence in the sample in favor or đ?&#x2018;Żđ?&#x2018;¨
Do not reject đ?&#x2018;Żđ?&#x;&#x17D; â&#x20AC;˘ insufficient evidence to support đ?&#x2018;Żđ?&#x2018;¨
• Select the significance level STEP 2: (α)
Significance level (α) • Typical level • 0.10, 0.05, 0.01 • Make decision • Statistically significant • did not occur by random chance • Not statistically significant • Show how likely the null hypothesis is true • Example, when α=0.10, there is 10% chance of rejecting the true null hypothesis
• Select the significance level STEP 2: (α)
Significance level (α) • Typical level • 0.10, 0.05, 0.01 • Make decision • Statistically significant • did not occur by random chance • Not statistically significant • Show how likely the null hypothesis is true • Example, when α=0.10, there is 10% chance of rejecting the true null hypothesis
STEP 3:
• Identify the test statistic
Test statistic • Population mean • (n>30) • (n<30) • Population proportion
Population mean
Test statistic • z-test • t-test
• The test statistic is the tool
• we use to decide whether or not to reject the null hypothesis.
• It is obtained by taking the observed value (the sample statistic)
• converting it into a standard score under the assumption that the null hypothesis is true.
• Large sample
• Use Z-test • When n ≥ 30, σ is KNOWN
• Small sample
• Use t-test • When n < 30, σ is UNKNOWN
Z-Test statistic
đ?&#x2018;Ľâ&#x2C6;&#x2019;đ?&#x153;&#x2021; đ?&#x2018;Ľâ&#x2C6;&#x2019;đ?&#x153;&#x2021; đ?&#x2018;§= = đ?&#x153;&#x17D; đ?&#x153;&#x17D;đ?&#x2018;Ľ đ?&#x2018;&#x203A; â&#x20AC;˘ Where
Population mean (n>30)
â&#x20AC;˘ đ?&#x2018;Ľ = sample mean â&#x20AC;˘ đ?&#x153;&#x2021; = â&#x201E;&#x17D;đ?&#x2018;Śđ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;Ąâ&#x201E;&#x17D;đ?&#x2018;&#x2019;đ?&#x2018; đ?&#x2018;&#x2013;đ?&#x2018; đ?&#x2018;&#x2019;đ?&#x2018;&#x2018; đ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;?đ?&#x2018;˘đ?&#x2018;&#x2122;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;đ?&#x2018;&#x203A; đ?&#x2018;&#x161;đ?&#x2018;&#x2019;đ?&#x2018;&#x17D;đ?&#x2018;&#x203A; â&#x20AC;˘ Ď&#x192; = population standard deviation â&#x20AC;˘ đ?&#x2018;&#x203A; = đ?&#x2018; đ?&#x2018;&#x17D;đ?&#x2018;&#x161;đ?&#x2018;?đ?&#x2018;&#x2122;đ?&#x2018;&#x2019; đ?&#x2018; đ?&#x2018;&#x2013;đ?&#x2018;§đ?&#x2018;&#x2019;
T-Test statistic
đ?&#x2018;Ľâ&#x2C6;&#x2019;đ?&#x153;&#x2021; đ?&#x2018;Ľâ&#x2C6;&#x2019;đ?&#x153;&#x2021; đ?&#x2018;Ą= = đ?&#x2018; đ?&#x2018; đ?&#x2018;Ľ đ?&#x2018;&#x203A; â&#x20AC;˘ Where
Population mean (n<30)
â&#x20AC;˘ đ?&#x2018;Ľ = sample mean â&#x20AC;˘ đ?&#x153;&#x2021; = â&#x201E;&#x17D;đ?&#x2018;Śđ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;Ąâ&#x201E;&#x17D;đ?&#x2018;&#x2019;đ?&#x2018; đ?&#x2018;&#x2013;đ?&#x2018; đ?&#x2018;&#x2019;đ?&#x2018;&#x2018; đ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;?đ?&#x2018;˘đ?&#x2018;&#x2122;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;đ?&#x2018;&#x203A; đ?&#x2018;&#x161;đ?&#x2018;&#x2019;đ?&#x2018;&#x17D;đ?&#x2018;&#x203A; â&#x20AC;˘ s = sample standard deviation â&#x20AC;˘ đ?&#x2018;&#x203A; = đ?&#x2018; đ?&#x2018;&#x17D;đ?&#x2018;&#x161;đ?&#x2018;?đ?&#x2018;&#x2122;đ?&#x2018;&#x2019; đ?&#x2018; đ?&#x2018;&#x2013;đ?&#x2018;§đ?&#x2018;&#x2019;
Z-Test statistic
Proportion • Binomial
𝑥−𝜇 𝑥 − 𝑛𝑝 𝑧= 𝑜𝑟 𝑧 = 𝜎 𝑛𝑝𝑞 • Where • 𝜇 = 𝑛𝑝 • 𝜎 = 𝑛𝑝𝑞 • 𝑝 = 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑝𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 • When • 𝑛𝑝 ≥ 5 𝑎𝑛𝑑 𝑛𝑞 ≤ 5
• Formulate the decision rule STEP 4:
Decision rule • z-test • t-test
Method Critical value
P-value
Differences Critical value
P-value
Critical value approach • A traditional approach • The critical value is the standard score that separates the rejection region (α) from the rest of a given curve.
P-value â&#x20AC;˘ The P-value for any given hypothesis test is the probability of getting a sample statistic at least as extreme as the observed value. â&#x20AC;˘ That is to say, it is the area to the left or right of the test statistic.
P-value
â&#x20AC;˘ Make statistical decision STEP 4:
Statistical decision