A researcher does a study of a mathematics program in a one-group pretest-posttest design. She gives the group of second graders a multiple-choice pretest. Unknown to her, the children had never taken a multiple-choice test. The children scored higher on the same test given at posttest. She argues that the children improved because of the mathematics program. That the children had never taken a multiple-choice test before and showed improvement at posttest suggests that one possible threat to her conclusion is a: