Key updates here are:
- In the 2-choice task (old/new), the output metrics are based on d' (target-lure, target-foil, lure-foil). There was a da reported in there that was vestigial and bogus. You can't get da without something like confidence ratings in there.
- The inclusion of a set of scenes used in Shauna's 2017 BBR paper comparing objects and scenes. Note that the lure bin file here and the ratings by lure bins are entirely fictitious. We've not run the hundred or so people needed to create bins. The file is just there because the code needs to have something in there.
PsychoPy study-test and continuous versions (beta)
I've been fully embracing Python and PsychoPy lately and have re-coded the stand-alone version into PsychoPy's Coder. Up on the new
GitHub page, I've placed both study-test and continuous versions. The study-test version forked off of the C++ code at 0.94 but incorporated the bug-fixes / updates in 0.96. Everything in the C++ version should be there now except for the touch-screen (Windows) responses. I'll be getting that in soon.
The continuous version was never made into the C++ stand-alone and existed here only in Matlab (PsychToolbox) code. It's PsychoPy code is very similar to the study-test, but relies on "order files" generated outside of the program. These tell the code what kind of stimulus to show when and build in all the constraints you may have in terms of the number of study-repeat pairs, the number of study-lure pairs, the distribution of their lags, etc. Sample Matlab code is included that will get ported to NumPy at some point.
Note, this code is beta-code. It's not been extensively tested. All bugs are mine, so let me know if you find any.
I've now placed the source code (C++ and PsychoPy), stimuli, samples / templates for instructions, etc. up on GitHub. Note, if you're looking to build the C++ code from scratch, you'll need to install things like wxWidgets, create a project in whatever development environment you're using, etc. I've got things going in MSVC 2010 and in XCode. But, I've yet to find anybody who really wants to do that, so ...
Craig