4 trials per participant, was thinking to give each participant a code of 1-3 (rating of consistency) but not sure where to go from here
It depends what you want to achieve. One possibility is to do an ordinal logistic regression analysis of the different answers, and look at the variance between trials, assuming that you have a sufficient and representative number of participants.
i have 66 participants. I aim to test the hypothesis that consistency wont be an overriding factor when sorting cards either taxonomically/thematically i.e. they might do a different sort each trial
It sounds like repeated measures, but a little hard to understand from your question. One thing you might consider though is how sensitive your measure is to change overtime. Having three scores of consistency seems like it may be too restricted range to have enough variance unless you expect your participants to make wide swings (e.g., go from a 1 to a 3). If the scale won't pick up on small changes because their are so few possible scores (i.e., 1-3), you may consider using/creating a scale with a larger range. Best of luck.
- Badges
- Science topic