The fidelity of the treatment was ensured by using the following procedures. All teachers
had at least 2 years of teaching experience and wanted to participate in the study voluntarily.
They were familiar with the teaching of modelling problems and received a 1-day training
session in the teaching method they were supposed to apply in the intervention. In each lesson,
at least one person from the research group was present. The members of the research group
administrated the tests and observed the implementation of the treatment. All lessons were
videotaped, and all students’ solutions were collected. Analyses of these materials revealed that time on tasks did not differ between the two conditions, and all students were asked to solve the problems as identified by the instruction manuals. However, in the multiple-solution
condition, students used the time for solving problems in two different ways. In the onesolution
condition, students discussed the solution longer than in the multiple-solution condition