I have to analyse data from an experiment in which participants performed two tasks. Each task had a different content and also a different "space condition". Content, space condition and order of presentation of content and condition were counter-balanced between participants. When participant finished the tasks, they answered a questionnaire about the difficulty of each one, containing a set of 18 questions (for each task). I ran a factor analysis for the 18 questions referred to task with space condition 1, showing that there was just a single component, and another factor analysis with the questions related to space condition 2, that also resulted in only a single component. My question is: Is it correct for me to run a repeated measures ANOVA comparing the two resulting factor scales (DV) as within subject IV ("space condition"), and with "combination of content and space condition" as between subjects IV? Thank you for your help.
|
|
||||
|
|
|
Factor analysis (FA) of your questionnaire tells you whether the questionnaire measures more than one thing. For example, a questionnaire on happiness might reveal a factor on job satisfaction and one on happiness of personal life Comparing the output of the two FAs would tell you whether the questionnaire's structure is different under the two different conditions. If that were the case, the questionnaire would measure different things under different conditions and would be useless to you to compare your subjects' reactions to the two conditions. But I doubt you can do a meaningful factor analysis with just 18 subjects. For self-created questionnaires you can either treat each item as a DV or the average of a subject's response over all items of the questionnaire. I'm afraid, I'm not expert enough to answer the more pressing part of your question as to how to structure the analysis itself. Looking for the repeated-measures tag should find you some more answers. |
|||
|
