I have a dataset containing information on 66 different subjects. Each
subject is presented with a series of 36 stimuli (including zero
magnitude) in random order. Their response to each stimulus is coded as
0 - not detected, 1-detected. Each subject is presumed to have a
"threshold" magnitude at which the stimulus can be detected. My goal is
to estimate that threshold (and its variance) for each subject.
I have been thinking of the threshold as the predicted stimulus magnitude
for which the probability of detection is 50%. My naive initial approach
was to run 66 separate logistic regression models. Is there a better
way? Thanks!
Dale Steele
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
