Yesterday I noticed that the total amount of time spent attending to two different areas of interest summed to a total greater than the amount of time they were present, which indicated a bug in my eye-data extraction program. The error, luckily, only affected areas of interest of arbitrary shapes (i.e., not circles or rectangles) and I have only used those once, and the data are not published. In the end, it looks like it will not affect my conclusions, but it was upsetting nonetheless.
I have three experiments with my new method under review in Learning & Behavior. Hopefully that paper will make it into print without too much additional effort and I will have it as a reference to the method in future papers.
The next paper I am preparing is on outcome tracking. It is based on some of the experiments I presented here. The paper will be showing that during conditioning people come to look on the screen for the appearance of the US spaceship, before it arrives. It will also provide an in-depth analysis of pupil data. In sum, I think the paper will be saying that A: Participants do learn “where” events occur in conditioning. B: When participants behavior is instrumental in producing an outcome they goal track- they look for that outcome. C: The context alone can support this outcome-seeking behavior. D: Pupils change diameter over conditioning- but are relatively useless as a measure of conditioning with the method. And finally, in total, although gaze direction does appear as a learned response, it is probably not as good of an index of general learning as is the behavioral keypress response.
When I get these data re-analyzed I’ll be posting them here to help organize my thoughts about how the various particular results fit with my conclusions above.