Share
Explore BrainMass

ANCOVA: Analyzing Therapy Effects

Using ANCOVA, compare stalking behaviors before therapy (stalk1) with stalking behaviors three months later after therapy (stalk2) by group (cognitive-behaviorial versus psychodynamic approach). Be sure to controls for how bad the behavior was to begin with (covariance).
Use APA format and SPSS.

Attachments

Solution Preview

See attached for tables and groups properly formatted.

ASSUMPTIONS:

Before conducting the analysis, I review the data for the assumptions for ANCOVA.

1. Normality. The skewness and kurtosis were above one for stalker2, meaning that it was not normally distributed. However, the general consensus is that violations of this assumption are not serious.

http://www.utexas.edu/courses/schwab/sw388r7_spring_2007/SolvingProblemsInSPSS/Solving%20Analysis%20of%20Covariance%20Problems.pdf

2. Linearity. A regression with stalker2 as the dependent variable and stalker1 as the independent variable resulted in a significant linear relationship (F=47.31, p<.001). So, the assumption of linearity is met.

3. Homogeneity of Variance. We will look at Levene's test to be sure this assumption is met. Our output shows that this assumption is violated, F= 7.189, p=.010. However, ANCOVA is robust to violations of the assumption of homogeneity of ...

Solution Summary

Your tutorial is 524 words, two boxplot groups and a reference. The assumptions are analyzed, the variables and hypothesis is presented and the analysis and results are discussed.

$2.19