kxkxk ANOVA GLM

Use  kxkxk_hw_data_133.sav

IVs are

  1. gender 1=male 2=female
  2. Total Social Support (TSS) – 1-low 2=high
  3. Stress -- 1=low 2=medium 3=high

DV is Depression (BDI – larger scores = report more depressive symptomology)

The purpose of the study was to examine the “buffering effect” and to compare the effect for women and men, There is a substantial direct relationship between stress and depression – those with greater stress usually show greater depression. The buffering effect hypothesizes that having good social support can “buffer” (moderate, inhibit, interfere with) this stress depression relationship. Said differently, among people with low social support, those with greater stress have greater depression, however among people with high social support, those with greater stress do not have greater depression. The study also explores whether the buffering effect is different for women and for men.

Perform each of the following steps, showing the output and answering all questions.

Getting Ready to Crunch

  • What is the focal IV in this study?
  • What IV do we want to determine if it moderates that focal IV?
  • So, we want to explore the simple effects of ______, for each level of ______, separately for _____ and for _____,

Initial Analysis

  • Get descriptive means. Be sure to set order or the variables in the UNIANOVA command to configure the Descriptive Statistics table the way you want it
  • Get means plots. Be sure to set the order of variables in the plot command to configure the plots the way you want
  • Getthe F-tests for the full 3-way design
  • Determine what effects are significant
  • Consider what lower-order effects we will need to check for descriptive/misleading patterns
  • Consider what lower-order effects are likely to be interesting – based on the aggregations involved

3-way Interaction

  • Get the cell means & follow-up analysesto examine the pairwise comparisons among the levels of Stress for each level of Social Support for each Gender
  • Usetables of “pointies” to describe the pattern of the 3-way using that selection

2-way Interaction Let’s explore the Stress * Gender 2-way

  • Get the Stress*Gender2-way interaction plot and of the corresponding simple Gender*Stress 2-way plots for each level of Social Support
  • Get the semi-marginal means and follow-up analyses to examine the pairwise comparisons among the levels of Stress for each Gender
  • Use a table of “pointies” to describe the pattern of the Stress*Gender 2-way interaction using that selection
  • Get the cell means & follow-up analyses to examine the pairwise comparisons among the levels of Stress for each Gender, at each level of Social Support
  • Use tables of “pointies” to check if the pattern of the Stress*Gender 2-way is descriptive or misleading for each Gender

Main Effects Let’s explore the main effect of Social Support

  • Get a plot of the Social Support main effect and of the corresponding simple effect of Social support for each level of Stress for each Gender
  • Getestimated marginal means & follow-up analyses to examine the pairwise comparions among the marginal means of the main effect of Social Support
  • Use a table of “pointies” to describe the pattern of the main effect of Social Support
  • Get the cell means & follow-up analyses to examine the pairwise comparisons among the levels of Social Support for each Gender, at each level of Stress
  • Use tables of “pointies” to check if the pattern of the main effect of Social Support is descriptive or misleading for all combinations of of Gender & Stress