Oh, I did get that part. The bit I didn't entirely follow when the authors had a longer discussion of different calculated phi values regarding the connection between the measured consensus effect and the participant's accuracy in the study. For one, I didn't recognize the term "phi" - Wikipedia implied that it might be the result of a chi-square test, which we did cover in the statistics 101 course I've taken, but it might have been too long ago as I'm not sure of how exactly that test applies in this case or how the phi value should be interpreted.