Chapter 10

I. t-test for two independent samples

A. Assumptions for t-test.

B. standard error for t-test is pooled from the individual sample variances

C. Formula for t-test

D. Degrees of freedom df

df = df₁ + df₂ = n₁ + n₂ - 2

II. Steps in hypothesis test using independent groups t

Same as with t for single sample

A. State hypotheses - now about relation of two means

Ho: m₁ = m₂ OR (preferably) m₁ - m₂ = 0

H₁: m₁ ­ m₂ OR (preferably) m₁ - m₂ ­ 0

B. Set alpha

C. Collect data and perform statistical test

D. Make decision about Ho:

E. Reporting result in APA format (only need to report type (directional or nondirectional) of hypothesis test if it is directional).

statistic (df) = value, p < level

t (15) = 2.22, p < .05

II. Testing the homogeneity of variance assumption

A. Variances must be equal for the independent groups t to be valid, especially when n's are different between groups.

B. One method of evaluating this assumption is Hartley’s F max test.

1. Compute F max

2. Find df - where df numerator = k (number of groups, 2 for t), and denominator = n -1 (assuming samples are the same size, use larger n when samples are not the same size)

3. Look up critical F in Table B.3

4. If F max is < 3 we usually don’t worry about it.