This study investigated the effects on the performance of different age group students of: (1) method of task presentation; (2) format of test questionnaire; and (3) the numerical content of the tasks. The sample consisted of 556 students from grades 7, 9 and 12. The tasks, used for assessing reasoning skills, were taken from the Video-Taped Group Test. Experiment one tested the effect of the method of task presentation (video-taped demonstrations versus paper-and-pencil tasks with ilustrations). Experiment two tested the effect of questionnaire format (multiple-choice versus short essay answers) and experiment three tested the effect of the numerical content on the responses of students from different age groups. Analysis of the data indicated that method of task presentation had effect only on young students (grade seven) performance, while numerical content had effect on the majority of the students, in all grades. The only students which were indifferent to numerical content changes were the formal reasoning students. The implication of these findings to science education and cognitive level assessment indicates the importance of concretization in the instructional process of young students and the importance of the mathematical aspects behind a scientific problem. (Author/PN)