Oh, also the hippocampus stuff discovered with single-cell recording is probably real? Mostly done since I stopped reading the field, but looks solid at a casual glance.
-
-
Replying to @Meaningness @literalbanana
I think it’s likely that *some* fMRI results are meaningful, but it’s clear most aren’t, and there’s no way for outsiders to tell which are which. I don’t know whether there are clueful people in the field who have good heuristics for discriminating. Maybe we’ll know in ten years
1 reply 1 retweet 8 likes -
Replying to @Meaningness @literalbanana
A good first-pass heuristic: How disputable is the fit between the concept and the measurement? The field is very hard to get a grip on, but single cell/low-level = good, fMRI/high-level = bad is not a good heuristic...
1 reply 0 retweets 1 like -
...in part because until recently single cell work has undervalued naturalistic experimental paradigms and neural population-coding. A lot of single-cell work is up for dispute.
2 replies 0 retweets 1 like -
Replying to @beausievers @literalbanana
Yes, to clarify, I didn’t mean that everything based on single-cell recording is robust. Rather that those particular findings (retinotopic maps, hippocampus) seem broadly believable to me (even though eg early vision turns out to be more complicated than thought 30 years ago).
1 reply 0 retweets 2 likes -
The criteria you gave for evaluating cognitive neuroscience seem like useful minimal requirements, but not nearly enough to give confidence. Are there fMRI results you consider solid? What further criteria do you apply to make you confident?
1 reply 0 retweets 2 likes -
Replying to @Meaningness @literalbanana
There are fMRI results I consider solid, with the caveat that I often disagree with others' interpretations of those results. That said, I think much of the work on vision is going to stick, including mid- and high-level vision, including object and action understanding. (contd)
1 reply 0 retweets 1 like -
A major development in this area is the use of multivariate pattern classification and representational similarity analysis to investigate the content of population-level neural codes, not just the magnitude of activity in a region
1 reply 0 retweets 1 like -
This way of thinking is feeding back to single-cell research, and we're seeing lots of nice studies showing similar results for single-cell and fMRI studies using similar experimental paradigms, which increases my credence
1 reply 1 retweet 2 likes -
I also think a lot of learning and memory research is likely to stick, and here I would include a fair amount of social neuroscience, especially frontostriatal reward learning work—which my on my (maybe idiosyncratic) interpretation is learning and memory research in disguise
2 replies 0 retweets 1 like
Thank you very much! These specifics are really helpful
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.