**Method** Participants were recruited to participate in a “decision making and attitudes survey” on Amazon Mechanical Turk (Mason & Watts, 2012). Participants were restricted to those who reported that they were American residents and have completed at least fifty hits with a hit acceptance ratio >95%. In Wave 1, 903 US participants were paid $0.70 for their time as a part of another experiment (Klein, et al., 2013). Participants completed a series of eleven experiments that closely replicated well known psychological phenomena (described below). Experiments were presented in a randomly assigned order and each consisted of two conditions to which participants were randomly assigned. Participants were randomly assigned to be recontacted and invited to participate in a research study either 2-3 days later , 1-2 weeks later, or a month later. All participants were paid $1 for their time for participating in the second study in which they completed the same eleven experiments again. The similarity of both study content and form was manipulated between the first and second experiment. By virtue of random assignment, some participants completed the same condition at time one and time two, while others were exposed to both conditions. Additionally, participants were randomly assigned to complete the study in either in a visually identical format, or on a different survey platform with its own distinct color scheme, fonts and layout (Qualtrics). This resulted in a 2 (Condition: Same versus different) X 2 (Platform: Same versus different) X 3 (Time; days, weeks or a month later) between-factors design. Wave 1 data was collected as a part of Klein et al., 2014. Methods and materials are posted here https://osf.io/abesq/