Participants were
randomly allocated to one of these conditions and order of testing
within sessions was counterbalanced so that half of the participants
completed a batch of questionnaires followed by a computer
task, while the other half had the order reversed. Based on an awareness
study by Roth et al. (2001) and a UEM study by Yeomans (1996),
effect sizes were calculated (Cohen’s d = 0.97 and 1.00, respectively),
and power analyses were conducted, showing that at least 18
participants were required per group to detect an effect (80% power;
p < 0.05).