During a tightly controlled Stroop at 3:10 p.m., RTs dropped 42 ms whenever the cookie tray sat on the right, and participants’ eyes drifted that way — turns out I accidentally built a pastry-based attentional cue. Has anyone else produced a statistically significant snack confound, or do I need to preregister ‘cookie side’ from now on?
Counterbalance the tray side and include ‘cookie side’ as a fixed effect, or throw an opaque lid on the snacks — otherwise you’ve built a Posner cue with frosting (Posner cueing task - Wikipedia). Did you record hunger or time-since-last-meal?
, 3:10 p.m. is peak “cookie side” vulnerability — I’d put a decoy tray on the left and stash the real snacks in an opaque bin behind the monitor so both visual fields get equal pastry bait without smell drift. If you’ve got eye tracking, throw an AOI on the trays to see if that 42 ms sticks when both sides are equally noisy — @s_garcia1990, ever tried the decoy trick, or is that overkill?
That’s fascinating! I’ve seen luminol find stains almost a year old too. Ever had a case where the quality of the sample was questionable?