Trying and struggling to understand this perspective. I’ve watched longer talks by Shiffrin and still don’t get it. https://featuredcontent.psychonomic.org/complexity-of-science-v-psprereg/ …pic.twitter.com/XAADkgdIg0
You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Yes, I think it has to do with differences in our subfields. I’d really like to get it, & know if he means it as broadly as he’s saying it.
Yes. I’m thinking of it now, on the bus. In social/emotion the models were mostly verbal, and the experiments created to fit standard linear statistical models, your t tests, regressions, anova, what have you.
Slight tangent, but the cool thing about cognitive modelling is anything can be a model. K-means b be a cognitive novel, logistic regression can be a cognitive model. Even a t-test can be a cognitive model. Depending on what you're modelling.
I’ve never done real formal modelling, but isn’t this a philosophical divide in parts of the field, where logistic regression would be rational model, a behavioral description of data and results, and a cognitive model would relate 1:1 with some kind of psychological construct?
Rational is such a loaded word and I do not think logistic regression is rational, it's a model, but not a rational one. So short answer: no. To decide on if a model is rational you need to formally define what you consider optimal behaviour for the task...
it's a minefield and that's partially why "rational" is a bit of a weasel word. I think a n appropriate question to ask, but cannot be answered here because I don't want to say more about the task/work, is what level of analysis (Marr, for example) the model operates at.
What I can say is I am using a logistic regression like I would use a neural network to model the participants (which BTW are already incredibly similar mathematically).
Those are not models of cognition and processes, but generoc testong models. You can use them on Beer, as well as psychology
It feels like a different beast than the verbal models. Though I think some of the verbal models could be turned into more computational models.
The model I used for my dissertation tried to specify the specific processing dimensions, mainly for visual stimuli. The tests and models were deeply co-dependent. I could then use generic stats to analyze, but some of the results were obvious
Others used custom transforms (but I gather these are common in various signal procesding).
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.