Students' difficulties in mathematics and science may stem from interference of irrelevant salient variables. We focus on the comparison of perimeters task, in which area is the irrelevant salient variable. A previous fMRI brain-imaging study related to the comparison of perimeters task suggested that increasing the level of salience of the relevant variable, perimeter, would increase participants' performance. The brain imaging results support and enrich previous behavioral data, and thus contribute to the development of our research rationale and intervention design. In the present study we increased the level of salience of the relevant variable by drawing the perimeters of the shapes as built from separate units, matchsticks (discrete mode of presentation), rather than drawing them continuously. Such increase in salience draws participants' attention to the perimeter. Thus strategies (such as moving of segments and/or counting them, or applying formal geometrical knowledge), that are regularly used when solving the comparison of perimeters task become more available. We explored whether the discrete mode would yield a higher success rate than the continuous mode and whether an intervention of first performing the discrete mode would improve students' success in a subsequent continuous mode. Findings show that success in the discrete mode was higher than in the continuous mode. Moreover, success in the continuous mode increased as a result of the intervention (i.e., when performed after discrete mode). The current study suggests that altering the mode or order of presentation can serve as an educational tool that may improve students' performance.