After reading the copies, some pages are mixed, for a test with multiple copies and two pages
First: thanks a lot for contributing to this project, which seems field tested and very flexible.
We have performed an experiment before the classes start and have found a small problem. Our first application is non-standard:
We want to pass a test for statistical purposes only, with no student identification. We used the options outsidebox,postcorrect,noshuffle.
The tests are read just fine, but in the final export, using the "ticked boxes" option, the rows are mixed. I mean, if test 1 had two pages 1:1 and 2:1, and test 2 1:2 and 2:2, then in the csv file I get a row (sometimes) with 1:1 and 2:2, and another row with 1:2 and 2:1.
For statistical purpuses, I get the marginals right, but the correlations are destroyed.
This happened with two pages from different pdf files.
How can I make sure that the pages from the same test stay together withouth identifying the student?I guess I can print different versions of the exam instead of making photocopies:
- would that work?
- is that the only choice?.
RE: After reading the copies, some pages are mixed, for a test with multiple copies and two pages - Added by Alexis Bienvenüe almost 5 years ago
- you can use the separate answer sheet mode to get a one-page answer sheet;
- you can print as many different versions of the subject as you need (even if these versions differ only by the ID boxes at the top).
RE: After reading the copies, some pages are mixed, for a test with multiple copies and two pages - Added by Juani Mela almost 5 years ago
Thanks a lot.
I was expecting the system to assign pages in the same pdf to the same test, but I guess these days it's not a big deal to use 200 versions of the test.