CPSY 501 Fall 2009 HW Assignment #3:
Mixed-Design ANOVA

Total: 20pts. Due: Fri 27Nov 09:00am, uploaded to myCourses

Dataset

Please download and use the HW3-BS_WEL.sav data file for this assignment. This data is an excerpt from Ben Schultz's thesis, "Weight Loss Self-Efficacy and Modeled Behaviour: Gaining Competence through Example."

APA Style (2pts)

Please format your answers as a complete APA-style paper. Be sure to include a title page, running headers, etc., and especially pay attention to formatting your statistical results in proper APA style, including reporting the effect sizes, df, test statistics, and p-values. Tables/figures must be formatted in proper APA style, as well. Write up your answers in a clear and well-formatted document (e.g., Word). Also, upload a properly labelled SPSS output file, and refer to the output to support your answer. You should upload two files to myCourses: your write-up (*.doc) and a SPSS output file (*.spv).

Mixed-Design ANOVA

Your data set consists of adults who participated in a workshop on exercise and health promotion supporting health weight maintenance. Some participants viewed a video that modelled success in exercise while others viewed a video of relaxing waves on the shoreline to support relaxation and stress management. Some participants had access to the videos for use at home during the follow-up period. The DV for this exercise is the Weight Efficacy Life-Style Questionnaire (WEL), tapping their current beliefs about their ability to lose weight. Conduct a Mixed-Design ANOVA to explore change in the workshop participants' WEL levels. For the purposes of this assignment, assume that the assumptions of interval-level measurement and normally distributed outcome variable have been met.

(1) Outliers (2pts)

Screen the data for problems with univariate outliers for the group as a whole (do not break down for gender or treatment group). Also perform basic data preparation / cleaning (e.g., check for missing data). Report any problems that you identify. However, do not remove / alter / transform the data here to adjust for outliers, or in any other part of this assignment.

(2) Sphericity and Equality of Covariance (2pts)

Check the assumptions of sphericity and equality of covariances. For each assumption, explain how you obtained your answer and report the appropriate statistics. If either assumption is violated, explain how this impacts our analysis.

(3) Time Effect and Interaction (6pts)

Do participants' WEL levels change significantly over time? How does this change relate to the "treatment effect" (interaction)? Interpret the interaction effect, illustrating with graphs as appropriate. Explain how you reached these conclusions, and report the appropriate statistical information. (Reminder: make sure you choose the procedure that matches your conclusion about the assumption of sphericity.)

(4) Gender as a Moderator (8pts)

Using the appropriate tests, identify and interpret all significant interactions with gender in this investigation. Conduct and report the tests, plot the appropriate graphs, and describe what that information tells us. How does this pattern of results change our interpretation of the treatment effect, if at all?