Millisecond Forums

Hi,
in a learned helplessness task I would like to insert a probabilistic feedback after a response is given by the subject, i.e. a feedback that is not contingent to the accuracy of the response. More precisely, the subject should receive a "punishment" feedback (in our case waiting for 12 sec until the next trial starts) in 75% of the trials and a "reward" feedback (waiting for only 2 sec until the next trial starts) on 25% on the trials but always irrespective of the correct or incorrect answer the subject has given.

Any idea on how I could  implement that in the script? Is there any option by using the select attribute? Or by using a counter? If yes, how would I concretely write it in the script? 

Thanks a lot for your help and looking forward to hopefully quick answers,
Michèle