2AFC data fitting

[Alex Burmester]

Hi Guys,

Congrats and thanks for developing this immensely useful tool!

I had a question about utilising Memtoolbox for 2AFC designs. Is it still possible to fit data from 2AFC to any of the models, or are some of the models (E.g., variable precision) only relevant to precision/error-based data.

If so, what form should I have my data in? Proportion correct for different stimulus increments?

Sorry if this is a stupid question, I couldn't find what I was after in the JOV article.

Cheers and thanks!

Alex B.

[Ronald van den Berg]

Hi Alex,

To my knowledge, the memtoolbox is meant for estimation data only. Models for 2AFC designs require a model of the decision stage (e.g. Bayesian observer, max model, etc), which can be combined with any model for the memory stage (slots-plus-average, variable-precision, etc). Each 2AFC task will require a different decision model, so it seems impractical to include 2AFC in a toolbox (but one could imagine a toolbox specifically for change detection tasks, for example). We have published several papers in which we combined 2AFC decision models with models of working memory. You might want to have a look at these:

Change detection:
* No evidence for an item limit in change detection, PLoS computational biology, 2013
* Probabilistic computation in human perception under variability in encoding precision, PloS one, 2012

Change localization
* Variability in encoding precision accounts for visual short-term memory limitations, PNAS, 2012

Visual search:
* Does precision decrease with set size? Journal of vision, 2012

Ronald

--
You received this message because you are subscribed to the Google Groups "MemToolbox users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to memtoolbox-use...@googlegroups.com.
To post to this group, send email to memtoolb...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/memtoolbox-users/28811a7e-caa9-493a-9e92-22c4c0d4c66e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.