I have a sample of 19 people who I measure for some factor. In step 2, I give them all some information and test again for the same measure. Step 2 is repeated 5 additional more times. So, in effect they are given additional information 6 times. Thus for each person I have 7 performance measures - one without information and six additional measures with information. I want to test if the performance improves with every additional piece of information. How can this be done?
The most common approach is probably a repeated-measures ANOVA with post-hoc tests. However, if you expect the change over time to be linear (or curvilinear) as a function of information provision (e.g., if the predictor is a count), you could also use a within-subjects regression approach, like hierarchical linear modeling.
Repeated-measure ANOVA is the options. You can also evaluate the relationship through this.
This comment is about the design more so than the analysis. It will be difficult to argue convincingly that the information you gave is responsible for any change in the outcome variable, as you (apparently) do not have a "no information" control group.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis