To manipulate AQ, in both the optimal and non-optimal conditions, we adapted the gist of the commonly used strong and weak arguments in favor of senior comprehensive exams provided by Petty and Cacioppo (1986). In the non-optimal condition, the content of the messages was taken directly from ML3's materials, whereas in the optimal condition, they were taken from Petty and Cacioppo (1986) with very minor modifications mostly for readability.4 The only substantive change was that the argument in the weak version linking failure to implement the exams to racial discrimination was changed to instead reference gender discrimination.