Presenting stimuli in different orders using responses.xls

I have an experiment with 4 different conditions assigned at the beginning of the experiment. Partway through the experiment, I want to present participants with 20 statements with varying ratios of congruent to incongruent information based on these conditions (i.e., 100% congruent, 75% congruent/25% incong, 50% cong/50% incong, 25% cong/75% incong).

I'm having trouble, though, figuring out how to present the 20 statements in a random order. I created a responses.xls file to do complex skips and calculations (e.g., to randomly assign which statements are congruent/incongruent). But if I make different versions of the MediaLab questionnaire in which the items come in different orders, and randomly assign Ps to these different orders, then the responses.xls file doesn't work right. For instance, if I have my ML questionnaire set so one condition has the statement order 3, 1, 4, 5, 2 and the items in my responses.xls file are listed as 1, 2, 3, 4, 5, then it won't work right because ML starts at #3 and doesn't know what to do when the next item in the questionnaire is #1 (because Excel already skipped #1 and #2 to get to the calculation stuff for #3).

If I have x different orders, I figured I could just have x different responses.xls files, one Excel file for each of the different questionnaire orders--but I think I can only have 1 responses.xls file per experiment, right?

The other option I thought of is that I could paste the different statement orders in different places in the responses.xls file, and just have ML look in the different places depending on which order is being used. For instance, order #1 (3, 1, 4, 5, 2) would be located in cells A1-C100, order #2 (4, 3, 2, 5, 1) would be located in cells A101-C200, etc. Is there a way to tell MediaLab where to look in the responses.xls file for the relevant skip/calculation instructions? Or does MediaLab always start at the beginning of the Excel file? Is there a way to set up a skip or something within Excel, so that ML will look at the right area?

I'm attaching the files I've been using if it will help--the relevant files are
"congruence manip - 100 - 1.que" and "responses.xls".

I would really appreciate any help!

Thank you!

Did our discussion regarding subject and condition IDs in the responses.xls file help you on this one at all?

Thank you--your suggestion helped!

What I ended up doing on the current version of the study was just to give participants all 40 sets of items and randomize the order in which they are presented since it isn't a major issue in the study if they get both sets of each question--it's just somewhat more boring for the participants. (It was more important to just start collecting data ASAP than to make the program perfect.)

I think for the actual study (this was just for pretesting), I'll take your suggestion and make the separate questionnaire files to have ML randomly choose which set of items each participant will see.

Sounds good, thanks for the update!