double-check distractor stimuli in tables
Closed this issue · 2 comments
For the stimulus table, distractors get their own row. The label for the distractors should be the label that is provided for that image in the experiment (when it is in target position). If the image never appears in target position, the label is either:
- the label given to the image in the experiment description (e.g., "the distractor was an image of a chair")
- if there is no obvious label provided in the experiment design, leave label blank
Fixed the ones I knew were broken, now we would need to go through all the datasets and check how this was handled. As this requires some digging into the designs (mostly: was every distractor image also present as a target at some point?) we could think about moving this into the general code review step instead of doing it on its own
Some errors could be build into the peekds validator - e.g. ensuring that each image path only appears once
Moved to checklist for review #119 , closing this one