How to estimate the systematic trial-by-trial variation in the drift rate with HDDM?

[Mei Yen Chen]

Hi there,

It was very nice to see this HDDM tool in the SfN-2012!

I've got a perceptual learning data set, and I was wondering if the HDDM will be able to recover the trial-by-trial improvement in the drift rate in my data. Has anyone try to fit HDDM this way? if so, could you tell me how to code the data for this estimation?

thank you very much for your reply in advance

Best,
Mei-Yen

[Thomas Wiecki]

Hi Mei-Yen,

On Sun, Oct 21, 2012 at 1:12 PM, Mei Yen Chen <yen...@gmail.com> wrote:

Hi there,

It was very nice to see this HDDM tool in the SfN-2012!

Thanks for stopping by!

I've got a perceptual learning data set, and I was wondering if the HDDM will be able to recover the trial-by-trial improvement in the drift rate in my data. Has anyone try to fit HDDM this way? if so, could you tell me how to code the data for this estimation?

I think that's a very interesting area of application for the DDM. We did experiment with this model quite a bit (in a completely different context) and it worked quite well. Essentially, you can define the nature of relationship between trial number and drift-rate (e.g. logarithmic) and estimate the parameters describing this relationship (i.e. the regression coefficients). Having said that, the model has not been ported to the newest HDDM version yet (and there is no prior release where it is available in a usable form). However, there are quite a few reasons for me to port this soon so I will get to this soon, hopefully within the next two weeks. So you'd have to wait until then.

Thomas

[Mei Yen Chen]

Hi Thomas,

thanks for your reply! I am looking forward to your new version! please keep me posted when you release this code! thanks in advance

Thomas於 2012年10月21日星期日UTC-5下午12時43分41秒寫道：