Visual inspection of a Data

1. Using visual inspection alone, which of these examples suggests an interaction exists in the data? To describe each of the examples, please hand draw a line figure and then decide if the lines suggest that an interaction may be present. a. b1 b2 a1 2 1 1.5 a2 7 6 6.5 4.5 3.5 b. b1 b2 a1 7 1 4 a2 5 3 4 6 2 b1 b2 a1 3 1 2 a2 7.5 6.5 7 5.25 3.75 c. d. b1 b2 a1 1 7 4 a2 5 3 4 3 5 e. f. b1 b2 a1 4.5 3 3.75 a2 4.5 3 3.75 4.5 2 b1 b2 a1 1 2 1.5 a2 7 6 6 4 4 g. h. b1 b2 a1 1 6 3.5 a2 3.5 3.5 3.5 a3 4 3 3.5 2.83 4.17 b1 b2 a1 2 5 3.5 a2 5 2 3.5 3.5 3.5 i. b1 b2 a1 6 1 3.5 a2 3.5 3.5 3.5 a3 4 3 3.5 4.5 2.5 j. b1 b2 b3 b4 a1 10 12 14 16 13 a2 8 10 12 14 11 9 11 13 15 j. b1 b2 b3 b4 a1 12 12 8 8 10 a2 8 8 12 12 10 10 10 10 10 2. Use the dataset Inspector_General_Example.sav. This dataset has two independent variables (behavior intervention and drug) and one dependent variable (frequency prosocial behavior). Please rename the variables in this dataset such that the independent variables and dependent variable reflect measured variables in your area of research interest. a. Now, run a 2-way analysis of variance with both descriptives and effect sizes included in the output and paste into your Word homework document. b. Please state each of your null and alternative hypotheses (there are 2 main effects and one interaction). c. Please interpret the results from the 2-way ANOVA based on significance level alone. 1. What is the interpretation for the first main effect? 2. What is the interpretation for the 2nd main effect? 3. Show a line graph and then interpret the interaction effect. d. Now, interpret the 3 results from the 2-way ANOVA in question #2 based on effect size magnitudes for each main effect and interaction. Remember to use rules of thumb for small, medium, and large effects. (hint: put the eta squared value into the G*Power software and use the Cohen’s f to interpret the effect sizes; or you can use the guidelines I provided in Lecture 3). 3. Now use the Two-way ANOVA Example2.sav dataset. Run a two way ANOVA using frequency of prosocial behavior as the dependent variable, and the other two variables as the independent variables. a. Paste the output into Word. b. Conduct pairwise comparisons between each of the Drug conditions using the LSD (no correction), Scheffe, and Tukey corrections for family-wise error rate. Include the output in the homework. c. Compare the three methods for adjusting the p-values in terms of the substantive results? Based on the output for 3b, does it matter in this example whether or not we control for family- wise error rate? Why or why not?