2
$\begingroup$

I am struggling with getting started using the glba package in R.

Often, the program will not fit the data at all. It immediately converges, but not really, and then instead of estimating it gives the starting values that I gave it, or it generates absurd values, such as negative non-decision times.

I tried to fit the data that is given as an example (bh08) and it worked great, however, with my own data that seemed superficially at least very similar (the RT distributions were in the same ballpark, and with a similar distribution shape) it did not work (converged immediately). I started with the suggested starting values, but also changed them.

Any smart advice?

(I guess RT should be in seconds and accuracy is a 1:0 variable. By the way, the program worked and did not stop when RT was in milliseconds, but the estimates were bizarre).

$\endgroup$
  • $\begingroup$ Could you provide a bit more info about what you are studying? What are your objectives, what do you expect to see, what have you tried and what are you using exactly? $\endgroup$ – Robin Kramer Apr 2 '17 at 13:58
  • $\begingroup$ Perhaps supply a reproducible example in your question and show how results are deviating from expectations: stackoverflow.com/questions/5963269/… $\endgroup$ – Jeromy Anglim Apr 3 '17 at 3:49
1
$\begingroup$

It appears as if the problem was caused by pre-processing using (the amazing) tidyverse package. When I (a) wrote the dataframe as a .csv file and then uploaded it, the LBA worked (in most subjects) nicely, converged quickly, and yielded sensible estimates.

Thank you!

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.