0

Possible Duplicate:
Why ANOVA/Regression results change when controlling for another variable

I have used SPSS to run two RM two-way ANOVA with the same data.

RM ANOVA #1 has two within subject variables A & B RM ANOVA #2 has two within subject variables A & C

Variable A is represented in both ANOVAs. What is strange, is when Variable A is reported as being significant in one ANOVA but not the other. Since the variable is the same, I expected regardless of the change in the second variable (B & C) that the first variable (A) should keep the same significance.

Is this possible or have I input something wrong in my data?

Community
  • 1
Barry
  • 1
  • 1
  • Provide more details on how you're specifying your ANOVA and what the variables are. – John Apr 01 '12 at 14:04
  • Thanks John, My RM ANOVAs were not measuring an assessment that was administered more than once over time but instead scores of test items were separated into categories. I'm testing the effect of two variables on the learning of vocabulary items. The First RM ANOVA is a 2 x 3 model and the second ANOVA is a 2 x 2. Variable A is the frequency of exposure to the vocabulary (high vs. low) and Variables B (two levels) & C (three levels) are other variables associated with the vocabulary items. – Barry Apr 01 '12 at 16:06
  • Also since there were not equal items in each of the categories (I had to enter the data in the 'long' format into SPSS), then I could not just enter the raw number of vocabulary items correct for each combination of the variables, instead I entered the % of items correct. – Barry Apr 01 '12 at 16:08
  • I have also double-checked all my data, it's all been entered correctly. – Barry Apr 01 '12 at 16:16

0 Answers0