There are a few ways you can do this. Since there are different subjects you want to take care not to carry over any values from the previous subject. The general idea is to create two temporary variables and RETAIN their values from one iteration of the data step to the next. One is for TIME and the other for TRIGGER_NUMBER. When captured this way the values from the previous observation become available for processing until they are changed in the current observation by an assignment statement. Now, you are all set to compare the values of these two variables from the current observation with the previous observation. This is presented below. Please note that I have added SUBJ. data oldtrigger ; input subj N Time Trigger_number ; cards ; 1 1 0 0 1 2 1.3 100 1 3 4.6 1 1 4 5 0 1 5 6 200 1 6 9.4 1 1 7 12 0 1 8 15 0 1 9 18.4 1 1 10 20 0 run ; data newtrigger ; set oldtrigger ; by subj ; retain lasttrig lasttime ; if first.subj then do ; lasttrig = . ; lasttime = . ; end ; else if trigger_number = 1 and lasttrig in (100,200) then do ; reacttime = time - lasttime ; end ; lasttime = time ; lasttrig = trigger_number ; drop last: ; run ; There is at least one other method to do the same (hint: LAG function). There are also some other ways but that would require an advanced understanding of the data step to supplant its default iteration with your own. Venky Chakravarthy ... View more

Sorry, I believe you need to contact Tech Support. I meant to send the same note for this question on reaction time and the question on regression. Essentially, this does not seem like a stored process question. Your best bet is to contact Tech Support for this. You have to have working program before you can turn that program into a stored process. To find out how to contact Tech Support, go to: http://support.sas.com/techsup/contact/index.html Good luck! cynthia ... View more