responses.xls is not working

We're having problems with our advanced feature file responses.xls.
Like in one of the examples we worked out a great excel file and wanted to test it in just one of five trials. The problem is, that it fills in the <variable> wording but that the skip to function doesn't seem to work.
Concretely, we want to skip only if three answers a like the desired ones (we already made it in the excel file and the report data showed that our idea works). Even though the calculation itself works it doesn't skip.
The problem seems to be between MediaLab and the Excelfile.

We thought of some reasons why the communication between th both doesn't work:

Is it necessary to put all the experiment variables + those that we created in the responses.xls file? Or would it be sufficient to have only the important ones?

Can you have more than 3 columns (e.g. to create a sum that we need in another calculation) or might this confuse MediaLab?

Do numbers in the variable names lead to confusion?

Maybe there's some other reasons but these came to mind although we already tried some different combinations and it didn't improve.

Thanks in advance

If the resulting [I]skipto[/I] value is included in the third column for the variable in question then it should work. A couple things to check for would be:

1. you have no blank rows in your [I]responses.xls[/I] file
2. try using lowercase variable names and do this both in your [I]que [/I]file and in your [I]xls[/I] file.

If neither of those two suggestions help, then please send me or post a [I]zipped[/I] copy of the experiment folder (at least the que and xls files) so I can try to replicate the problem here.

Thanks for your answer but we already found the solution.
We changed the calculation and furthermore put the result of the calculation in the third column of an experimental variable instead of the calculation variable (I think this might have been the main problem before) so that it works perfectly now! But thanks, anyway.

Another question: Is it normal that with responses.xls the change of each display needs a lot of time (1-2 sec.)? We've got a rather long experiment so it's quite disturbing.

I have seen delays like that. And, yes, I do believe it has been in cases where there have been many variables involved in the study. Not really a way around this aside from faster computers or more efficient program code. Before we release the next version we'll take a look and see if there is anything we can do to speed this up. Also, I assume you are running the study locally on each machine and not across a network? I mention that because that too can slow things down under heavy loads.