Hi guys,
Is there a limit to the number of questions that you can ask in an experiment? We're experiencing a strange problem with one of our ML2008 experiments. We set up a study that involves a fairly large number of questions (a bunch of responses for each of four replicates, plus two scales). We found that ML would simply stop, presenting a blank screen pretty far into the study (on the second scale), requiring task mgr to exit, and no data would be recorded for the participant.
Thinking that it might be something specific to the scale where it stopped, we reordered the two scales. It still stopped, but now in the other scale (at about the same total response number). Thinking that there might be something related to the number of questions, we ran the scale by itself, and it worked just fine. Then we ran copied the scale about ten times in the same study - it crashed at about item 156. Thinking that it might have something to do with an internal algorithm that renames variables to create unique variables, we renamed all vars to be n1 to n300. It still dies at 156. The strange thing is, we have other studies that are longer than 150 items. Would the type of item matter? (These are all scale responses on a 1-5 scale.) We're perplexed...can you help us out?
Thanks,
Alan