One possibility is that students did not take the pair-share opportunities seriously. If they regarded them as simply breaks from the lecture routine and a chance to chat with fellow students, rather than as learning experiences, or aids to recall, they may not have paid them much attention. They did not bother to store in memory what they thought was casual conversation. Dorr, Dill, Anderson, and Heppner (1996) asked their students to write out their thoughts on a worksheet before they shared them with another student in class. The worksheet may have increased the students' awareness of the seriousness of the task. It also gave students a written record of their thoughts to keep with class notes, and may be the critical element in the success of the pair-share technique.

A second possibility involves what Dorr, Dill, Anderson, and Heppner (1996) point out is a potential pitfall of the pair-share technique: belief perseverance. Students who held mistaken beliefs about what was covered in lecture, and those who had misunderstood, might further elaborate their error in discussion with their peers—the blind leading the blind, as it were. In such a situation, the pair-share technique might actually weaken exam performance, particularly if students with mistaken beliefs can convert their peers to an erroneous way of thinking. Their pair-share experiences are recalled but do not help on exams.

Further study is needed to clarify the differing results with the use of pair-share. There may be other ways to encourage self-reference that would work more reliably. In the meantime, instructor self-disclosure is clearly useful to students' understanding of concepts presented in my class lectures. Gossip is memorable.