The aim of this paper is to compare the effect of activity-based teaching on remedying probability-related misconceptions of students at different grades. Thus, a cross-sectional/age study was conducted with a total of 74 students in 6th-8th grades. Experimental instructions were given to all the groups three times/ week, 40 min/session, for 2 weeks. Students’ progress was examined by pre-test and post-test measurements. The results of the analysis showed that, as a result of the intervention, all graders’ post-test scores regarding all the concepts (PC: Probability Comparison, E: Equiprobability and R: Representativeness) showed a significant increase when compared to pre-test scores. It was found out that this increase did not create a significant difference based on age in PC concept, but that in 8th grade students, it showed a significant difference in E and R concepts compared to 6th graders. On the other hand, it was also assessed that the increases observed between 7th and 8th graders with regard to E and R concepts were not significant. In summary, the implemented intervention can be suggested to have different effects depending on age and the concept.