In an experiment, teachers' judgments are used to measure the amount of misbehavior among students. The teachers are more likely to judge a given behavior as misbehavior at the posttest than they are at the pretest. This example most clearly illustrates which threat to the internal validity of the experiment?