Randomization and Skip Patterns

I have followed the instructions in the MediaLab manual to randomly
assign subjects to conditions in an experiment of mine, but it doesn't seem to work how I anticipated. Let me explain what I've done.

In my experiment, I gave each condition the same WGR value. Moreover, each of these conditions has the same SKIP TO value. I expected MediaLab to randomly pick one of these conditions, and then have it skip to the question I indicated. But instead, the three conditions are randomly presented before skipping to this question.

Can you offer any advice on this?

Efren

Are you using a multiple response item type? I'm wondering if maybe the question is of a type that doesn't allow for the usual skip pattern functionality. If not, please send or post a copy of your que and I'll take a look at it.

The items that are not skipping are individual instruction items, which serve as experimental treatments. These items are then supposed to skip to a scale response. I have included my questionnaire. The items are #43 - 49.

Thank you.

Can you try this shortened version and tell me if you get the same problem. As far as I can tell, it randomizes as it should.

Thank you for your help. I have been away on business and am now returning to my experiment.

After I reported this issue to you, I re-installed an older version of MediaLab on my computer (v. 2004). When I did that, the randomization/skip patterns worked as expected. Moreover, when I ran the short file you sent me, the randomization/skip patterns also work as expected. Is this a software issue? Or did I do something incorrectly?

My questionnaire was programmed using v.2004. Is this going to be a problem if I administer the experiment with laptops that have a newer version of this software, or should it transfer over seamlessly?

Thank you,
Efren

We're using v2006 here and it seemed to run fine. For a comment on compatibility, see:

[URL]http://www.empirisoft.com/support/showthread.php?t=72[/URL]

My first suggestion would be to try running it on any other machine that has v2006 installed. You could use a free trial license if you simply want to test it.