I'm looking at differences between two groups (one who played the game, one who used a different app) -- so I'll be looking at differences in the count of those action sequences, and also differences in counts of the different themes in their short answers. I have a good understanding of what statistical tests I will be doing (that's not the question), I just need a better understanding of which data to actually use for the tests... whether or not I use data from rater-1, data from rater-2, or some sort of consensus/combination of both. Sorry if I was not clear earlier!

My study involves students playing an educational game... For the continuous variable, the raters are counting the number of times players perform a certain sequence of actions (from video footage). For the categorical variable, I have written phrases that are coded by theme (4 themes total). Thanks for any advice!