“…5 While comparison with the control group is, by definition, impossible in the context of the writing exercises, comparison was possible in various analyses of assessment tasks the students were required to complete throughout the year. The one criterion on which there was noticeable divergence between experimental and control after the beginning of the writing intervention was that of understanding 6 (see also [15,28,35]). In a number of assessment tasks throughout the year, students' success at completing problem-solving questions was measured on a number of variables (such as raw grade obtained, indication of understanding, apparent metacognitive control, flexibility of solution), with understanding of the problem showing noticeable improvement for the experimental group compared to the control group.…”