PSYCH 100A Lecture Notes - Lecture 11: Null Hypothesis, Variance, Analysis Of Variance
Document Summary
Tests three or more group mean difference by comparing two estimate of population variance on x. Affected if the null is true or not. The more the spread of the group mean differences, the likely the null hypothesis is false. The group mean is usually clustered to the grand mean score if the null hypothesis is true. One factor anova is the appropriate analysis for research questions with one categorical independent variable and one numeric (continuous) dependent variable. Mean square (ms) = s^2 = ss (sum of squares) / df. Extension of the two-sample independent groups t-test to situations where there are more than two groups. Three groups: a placebo group, a low dose of the drug, and a moderate dose of the drug. Grand mean= just add up all the scores and divide by total number of samples. Between group variance, compare the group means to the grand mean. Degrees of freedom= total of n - amount of groups.
