exercise 4 a-c

FURTHER EXERCISES These further exercises are challenging, but important. They are designed to facilitate deep thinking about the formulae you are working with, not just mindlessly plugging numbers in. Exercise 4 Quick questions a. When do we use ANOVA? (refer to the type of IV and DV, and number of categories) b. Why is it not recommended to use ITT for 3+ groups? (see page 4 of lecture slides) c. Sometimes F = t'. In what circumstance does this relationship apply? -7 - Exercise 5 True or false? a. As 'variance between' increases, F-score decreases. b. If there is a lot of 'noise', the F-score is likely to be significant. c. Variance between = Mean Square Between = numerator variance = "effect" d. Variance within = Mean Square Within = denominator variance = "noise" e. We need to know the value of Mean Square Between to calculate differences between groups when using Bonferroni's post-hoc testing. f. To perform an ANOVA/post-hoc analysis, each category must have equal number of subjects. g. The alternate hypothesis of an ANOVA test is that all the groups are significantly different from one another. Exercise 6 Similarities and differences For the following tests, list the elements required to find the critical score from the distribution tables. Regarding df, ensure that you specifically list out the unique formulae for each test. One-sample t-test, paired t-test, ITT, ANOVA, Bonferroni, Fisher's LSD