The research question of this study was, ‘how is the statistical reasoning ability among the tenth-grade students in secondary schools?’ In the variable map of Rasch measurement model analysis, we notice that most of the students are located below the zero logits. Lower logits manifests students with lower ability.This proved that the students’ statistical reasoning ability was absolutely poor. On the other hand, the measurement of person reliability using Rasch measurement model was poor at only 0.43 and person separation was also very low, i.e. 0.86. Students were not differentiated well in this study due to inadequate items and narrow ability range. Consequently, to gain good person reliability, it is recommended that more items should be added into the test and students’ ability range needs to be broadened. In addition, for the person analysis report, only five out of 412 students belonged to good fitting type. There were 20 misfitting students where 10 of them were overfitting; ten of them had negative value of point-measure correlation which had impinged on the results. It is suggests that these ten students are to be eliminated from this study.