Based on a criterion test 27 fifth and sixth grade students were selected who could not control variables. Three groups were formed. The control group received no training; a second group received training relying upon external reinforcement, and the third group, training designed to induce internal cognitive conflict. The groups receiving training did only slightly better than the control group on the posttest. On the retention test one month later, all three groups had almost identical scores. The mean combined retention test score was found to be significantly greater than the pretest score. The improvement of the control and conflict groups supports Smedslund's8 finding that external reinforcement may not always be necessary for behavioral changes to occur.
A further analysis of test scores led to the classification of students according to Piaget's developmental stages. During the course of the study a significant number of students improved in their stage placement. Based on retention test performance nearly half of the students having a mean age 11.8 years could be classed in late formal operations with regard to their ability to control variables. Inhelder and Piaget12 reported that this level was not normally reached until 14 or 15 years of age. In conclusion, the results support the possibility that while development of the ability to control variables can be accelerated by training at the fifth and sixth grade level some children are surprisingly unaffected regardless of whether the training is based on external reinforcement or cognitive conflict procedures.