Hi!
I am running an experiment where my “participants” are performing the same type of task 6 days in a row. My descriptive stats show that performance improves across trials.
I have eight participants, 6 trials run on each one. The performance is measured in how fast they are at solving the task – it can take anywhere from 0 to 3600 seconds.
The data is NOT normally distributed. Most tend to score either very low (0-500 seconds) or fail at the task (3600 seconds).
I want to analyze whether there is a difference between participants, and also whether there is a significant change across trials (although the descriptive stats clearly show that there is one).
Am I correct in thinking I need to
1. Transform the data to make it normally distributed
2. Run a repeated measures ANOVA (trial x participant x score) with trial as the repeated measure
…or is there a better way to analyze the data? Perhaps a non-parametric test that would not make it necessary to transform the data?