The hypotheses that interpersonal learning is more effective than individual learning in nonlinear tasks but that students learn relatively less from their teachers when the teachers have to teach a complex nonlinear rule than when they have to teach a simple linear rule was tested in an experiment following the social judgment theory interpersonal learning paradigm. A 2 x 3 x 4 factorial design was employed: (learning tasks: linear vs. nonlinear) by (learning conditions: teachers, who already knew the task, vs. learners, who had to learn the task from the teachers, vs. individual learners, who had to learn the task from outcome feedback from the task) by (blocks of trials). The results supported both hypotheses; the subjects learned the nonlinear task much better in the interpersonal learning conditions than in the individual learning conditions, but for the linear task, there was very little difference between the two learning conditions. Results also indicated that learners approached the same level of performance as their teachers at a slower rate in the nonlinear conditions than in the linear conditions. (Author/RD)