-1
$\begingroup$

Similar to the question "How to use a QUEST staircase in 2-AFC”. I also have a question about using QUEST. In fact, I too tried to find an answer in

Psychophysics: A Practical Introduction by Kingdom & Prins; Experimental Design by Cunningham & Wallraven; Visual Psychophysics by Lu & Dosher.

Although I found a script of QUEST, it is just for the detection task, where the threshold is always the upper threshold. I want to study also the lower threshold in the discrimination task. Is anyone can give me some hints?

$\endgroup$
  • 1
    $\begingroup$ Welcome. What do you mean lower threshold? I have never heard of an upper or lower threshold; the threshold is a single estimate; it is the threshold, not a threshold. Are you asking whether QUEST can be deployed for a JND task? $\endgroup$ – AliceD Jul 31 '18 at 8:31
  • $\begingroup$ Yes, I want to use QUEST in a JND task. The upper threshold means that standard stimulus + threshold > standard stimulus, the lower threshold means that standard stimulus - threshold < standard stimulus. For example, in a JND task, the JND = (25% threshold + 75% threshold) /2, however, how can I get both 25% and 75% threshold? $\endgroup$ – Francis Aug 1 '18 at 11:21
1
$\begingroup$

QUEST is one method of measuring thresholds. There are lots of other procedures including method of constant stimuli, modified up-down staircases, PEST, and ZEST. Each has their own set of requirements and own set of assumptions. If, as you say in the comments, you want to define the JND as the (25% threshold + 75% threshold)/2, then QUEST is probably not the procedure you want to use. The QUEST procedure is designed to optimally find the parameters of a psychometric function. While these parameters can give you the information you want, it is not necessarily the optimum procedure.

As it sounds like you are just learning, you may be better off with the method of constant stimuli. You present stimuli at a variety of levels, record if the answer was correct/incorrect, fit a psychometric function and then read off the values.

| improve this answer | |
$\endgroup$
  • $\begingroup$ Thank you very much for you answer! In fact, I have to measure the threshold corresponding to a certain proportion correct (e.g., 75%) first so that I can design constant stimuli for each individual. Anyway, I decide to measure the 75% threshold and use it as the step between constant stimuli. $\endgroup$ – Francis Aug 3 '18 at 15:46

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.