Dear TIPS folk,
I'm thinking on how to best implement an experiment - any thoughts?
I would very much appreciate any input...
I'd like to find out the best stimulus rate (or the inter-stimulus
interval) for a simple visual suprathreshold detection task with a
very high stimulus probability (0.75). About 100 stimuli are
presented for each eye of an observer. (Don't laugh: these tasks
are very common in clinical work). The ISI presently used is between
1000 and 1500 ms
I feel that this is much too slow for some fast observers, who then
get bored and loose motivation / concentration.
Ideally, I would like to express the stimulus rate in terms of
multiples of reaction time (or better: response time).
My current setup starts at an ISI of 1000 ms. The observer is then
asked whether they'd prefer the task to run faster or slower.
Depending on this subjective response, the ISI is made shorter or by
100 ms, and a new run (100 stim) is started. This goes on until 3
reversals (faster--slower, slower--faster ) have been obtained.
Problems with this:
1.) Trials go on for very long, and there are pronounced practice
and fatigue effects
2.) The reversal points depend entirely on the subjective judgment
of the subject.
Any ideas how I the design could be improved??
Paul
Research Assistant
Royal Eye Hospital, Manchester, UK
[EMAIL PROTECTED]
