Varying problem effort and problem completion rates : an investigation of the interspersal procedure and student assignment choice

The discrete task completion hypothesis suggests that, when given assignments comprised of multiple discrete tasks, completed discrete tasks are reinforcing events (Skinner, 2002). The current experiment consists of two studies investigating the interaction of relative response effort and relative problem completion rates on student assignment choice and ranking behavior. College students participated in Experiment One. In order to enhance the educational validity of the findings from Experiment One, Experiment Two was conducted with sixth-grade students. In Experiment One, college students were exposed to two pairs of mathematics assignments. Assignment Pair A included a high effort assignment containing 18 long 3- digit x 2-digit multiplication problems with all numerals in each problem being equal to or greater than four and a moderate effort assignment that contained nine long problems and nine interspersed moderate 3-digit x 2-digit problems with numerals less than four. Assignment Pair B contained similar assignment sheets, the exception being that the high effort assignment contained six additional I-digit x I-digit problems interspersed following every third 3-digit x 2-digit problem. Analysis of Assignment Set One revealed that students overwhelmingly preferred the moderate effort assignment. Analysis of interaction effects showed that when additional brief problems were added to the high effort assignments, the proportion of students who chose the high effort assignment for homework and ranked it as being less difficult, time consuming, and effortful increased significantly. Results support previous research on effort and the hypothesis that a completed discrete problem may serve as a reinforcing event. Experiment Two was a replication of Experiment One. However, the educational validity was enhanced, as participants were sixth-grade students who were exposed to assignment pairs identical to those of Experiment One. Results were similar to those found in Experiment One. Results from both experiments showed that students were more likely to choose assignments that required less effort to complete. However, when high effort assignments were altered by interspersing additional brief problems, the probability of students choosing the high effort assignment for homework increased significantly. These results support the discrete task completion hypothesis and suggest that educators can increase the probability of student engaging in assigned work by giving them more work. Additionally, logistic regression analysis yielded models of choice behavior similar to those found in previous studies of the matching law (e.g., Baum, 1974). This novel analysis of group data provided additional support for the discrete task completion hypothesis.