After reading the article by Enochson and Culbertson (2015), I found out that reaction times elicited via self-paced reading tasks should be analyzed based on the residual reading times if the interest regions are not equal in length or number of words. Up until now all I read was that if necessary, reading times needed to be log transformed. The question is whether analyzing residual reaction times is a standard for self-paced reading studies or not. I would be grateful if you could share some research experience / insights regarding this matter. I am a Master's student and I am still trying to figure out what is best to do with the data that I will gather via the series of online experiments that I have planned. In my case, the the critical and spillover regions differ in word length and this is what makes me think I should opt for looking at residual reading times, but from what I have read, the common thing is to simplay apply linear mixed-effects models on the raw or log-transformed reading times and leave it at that.
Your help would be very much appreciated!
Enochson, K., & Culbertson, J. (2015). Collecting psycholinguistic response time data using Amazon mechanical turk. PLOS ONE, 10(3), e0116946. https://doi.org/10.1371/journal.pone.0116946