2

I am doing a pre- and post-test intervention. I have two groups, each doing a different intervention. I will test the success of the interventions using three outcome measures.

In another way to look at it, one group is the control group. One of the outcome measures is a check to make sure that the control group's treatment is effective. The other two tests are the ones specific to the success or failure of the intervention. The hypothesis is that Group 1 will improve on test A and test B, but Group 2 will not; Group 2 will improve on test C, but Group 1 will not. I need test C to ensure that Group 2's intervention was effective, as it will enhance the confidence in the results of Group 1 on test A and B.

I am assuming this is an ANOVA. Is that correct? Is it correct to call it a 2x2x3 ANOVA?

gung - Reinstate Monica
  • 145,122
BTM
  • 21
  • 2

1 Answers1

2

Are the three outcomes measures on the same metric? If not, then you would need to standardise the outcome measure to put them on the same metric in odrer to include multiple outcomes in a 2 by 2 by 3 mixed ANOVA.

The three-way interaction would then test whether there is a differential effect of the intervention across outcomes.

This question discusses general issues related to studying single outcomes in pre-post treatment-control designs.

Community
  • 1
Jeromy Anglim
  • 44,984