College students took standard Introductory Psychology tests. The rate
at which they answered questions and the accuracy with which the ques
tions were answered changed systematically across test sessions. The p
attern of the change depended on the consequences of the behavior. Whe
n the score on the test counted toward the course grade, the percentag
e of items answered correctly and the rate at which items were answere
d either increased across the test or increased and then decreased. Wh
en the test score did not count, the percentage of items answered corr
ectly increased and then decreased across the session for those who pe
rformed relatively well. It decreased for subjects doing poorly. The r
ate at which items were answered did not change systematically. These
results resemble those for nonhuman subjects responding on standard op
erant procedures. In both cases, responding changed systematically acr
oss sessions and the pattern of this change depended on the consequenc
es of the behavior. Finding such similarities suggests that systematic
changes in responding across experimental sessions may be a relativel
y general characteristic of animal behavior that may have important im
plications for theory and research.