Preprint reviews by Jessica McFadyen

Category learning biases sensory representations in human visual cortex.

Edward F. Ester, Thomas C. Sprague, John T. Serences

Review posted on 06th September 2017

We discussed this paper in our lab meeting and so I thought I'd share some of the comments in case they're helpful:

* It might have been valuable to do scanning beforehand so that there were neural measures before people became experts at orientation discrimination.

* In Figure 4, the responses for two and three steps look noisier, likely a result of smearing across subjects. Would be interesting to see the results of individuals, ideally showing the tuning functions for each person.
- What would the results look like if, instead of categorising the EEG results into 0-125ms and 126-250ms time bins, it had been looked at continuously over time?
* Q: Is this persuasive evidence of “lower-level sensory” processing vs. “higher-level” processing?
- A: For the fMRI, might be more persuasive if the encoding results were just in the early visual cortex (but then perhaps this is reverse inference)
- A: The EEG and fMRI experiments each show that the effect is early temporally and hierarchically, which is persuasive when taken together.
- A: Maybe the question isn’t whether it’s “lower” or “higher” level processing, but rather that it’s both and how does this depend on exposure (training) and what the task is?

Thanks for sharing this work on bioRxiv! Looking forward to seeing it published, very interesting work.

show less