Transcript 11/24/2015
Psychology 202a Advanced Psychological Statistics November 24, 2015 The plan for today • • • • • • A priori contrasts in SAS Orthogonal contrasts Contrast coding The Eysenck ANOVA example Helmert contrasts Introduction to power A priori contrasts • A contrast is a question about a linear combination of means. • Example: H0 : Massed Spaced 2 None 0 • Shorthand notation: 1/2 1/2 -1 • Equivalent: 1 1 -2 • Another question that might interest us is 1 -1 0. Contrasts (continued) • Once a contrast is specified, its sum of squares is calculated: k SS contrast c i M i i 1k 2 2 c i n i 1 i • Contrasts always have 1 df, so the sum of squares is a mean square. • Division by the error mean square provides an F statistic that tests the contrast. Contrasts (continued) • Illustration in SAS. • Any set of contrasts defined in advance may be tested, dividing the alpha among them. • However, this particular set has a special property: orthogonality. • If the contrasts are orthogonal and specified in advance, there is no need for an adjustment to alpha. Checking for orthogonality • Multiply the corresponding coefficients of each pair of contrasts. • If the products sum to zero, the pair is orthogonal. • Here, we are considering (1, 1, -2) and (1, -1, 0). • (1×1) + (1×-1) + (-2×0) = 0, so the pair is orthogonal. Why is orthogonality special? • Contrast coding • Illustration in SAS • So orthogonal contrasts divide the model sum of squares into exhaustive and mutually exclusive partitions. • A more complicated example (Eysenck memory experiment) Introducing power • In the world of hypothesis testing, one of two things is true: – The null hypothesis may be true; or – The null hypothesis may be false. • In the world of hypothesis testing, one of two outcomes will occur: – The null hypothesis may be rejected; or – The null hypothesis may be retained. Consider that in tabular form: H0 True H0 Rejected H0 Retained H0 False Consider that in tabular form: H0 True H0 Rejected H0 Retained H0 False Great! Consider that in tabular form: H0 True H0 Rejected H0 Retained H0 False Great! No problem. Consider that in tabular form: H0 True H0 False H0 Rejected Type I error Great! H0 Retained No problem Consider that in tabular form: H0 True H0 False H0 Rejected Type I error Great! H0 Retained No problem Type II error Consider that in tabular form: H0 True H0 False H0 Rejected Type I error (p = a) Great! H0 Retained No problem Type II error Consider that in tabular form: H0 Rejected H0 Retained H0 True H0 False Type I error (p = a) Great! No problem Type II error (p = b) What is power? • In that scenario, power = 1 – b. • In other words, power is the probability that we will avoid a Type II error, given that the null hypothesis is actually false.