I don't know if you would consider it a parallel version, but the Stanford-Binet has a matrices test that is modelled on the Raven. Downside: you have to get the full test to get the matrices subtest and I'm not sure that the ceiling on the test would be high enough for your purposes, given that it was modelled on the Standard Progressive Matrices and not the Advanced version. If you are looking for a very challenging and possibly frustrating test, have you considered The Remote Associates Test (available only in English as far as I know)?

Dear Rhianon Allen, thank you very much for your kind and punctual reply.

I saw that the Stanford-Binet test is also designed for children, but my sample would be young adults so it should be adapted for this too.

I never heard about The Remote Associates Test, I will definitely take a deeper look at it because we are actually looking for a test that measures frustration tolerance, and "how easily a person gives up when they can't find a solution".

A persistence variable (= how long a person is able to 'stay there' trying to solve a pattern), measurable (we thought) with very demanding or impossible tests in which the solution is not there and so we can calculate the time that people decide to dedicate to solving the scheme before giving up on it.

I needed two parallel versions in difficulty to administer them pre- and post- experimental exposure. Not finding them, I was thinking of building them from already existing shortened versions of the APMs and creating them with algorithms/softwares that I found in other papers.

I also found a test of matrices similar to Raven, it's called Heidelberg Figural Matrices Test (HeiQ).. it's very recent but a parallel form has already been created.

I apologize for being long, thank you again!

Tomas Scheimovits