[edited from support email]

I want participants will get feeback about how they respond to stimuli (i.e., whether their response is correct/incorrect). I can using the jumping function and get it to display either an incorrect or correct message depending on their response. But what I'd like to do is to have the system select at random from a number of alternative messages to display. For example, rather than always displaying "correct", I want to display "great", "wonderful", "awesome", etc., and I want to display them at random.
First I tried to "rand(~BAD,~AWFUL, etc.)" in the stim field on trial
997 (which is the trial that is skipped to when incorrect response is made). I also tried making a text file of the incorrect messages and call it with $badfb.txt (also placed in the stim field on trial 997).

But in both cases, what happens is that the FIRST time a word is selected from the list, it is selected at random. So each participant will get a randomly selected positive or negative feedback feedback word. But then it is repeated every time thereafter. So it does not randomly select a word from the list each time the list is called. So participant #1 may get BAD or AWFUL or TERRIBLE at random the first time they give an incorrect response, but each and every time they give an incorrect response thereafter, they're going to get the same feedback word.

Am I missing something in the manual that describes how to do this specific thing? A rather tedious workaround would be to create 2 skip trials for each experimental trial (e.g., make a skip to 997 if response is correct and 998 if incorrect for trial #1, and a skip to 999 if correct and 1000 if incorrect for trial #2, and so forth). But that would be painful.

Any help would be appreciated!