An alternative for mathematical word problems is proposed.
An instrument for investigating problem representation is made available.
The instrument measures students’ performance after changing representation.
In the instrument a randomized controlled trial is embedded.