Participants took a survey rating their opinions on political statements and the certainty of their responses. A portion of the survey was shown back to the participants with half of the responses secretly manipulated. Participants were either placed under the acknowledge condition (participants simply have to acknowledge the manipulated response on screen) or the confabulation condition (participants had to acknowledge the manipulated survey result and provide argumentative reasoning to the survey response). Participants then had to fill out another questionnaire including all political statements from the previous one along with 6 new statements and then took the questionnaire again one week later. The researchers hypothesized that the amount of confabulation a participant engages in when justifying a manipulated survey response will increase the self-induced attitude change, as well as persistence over time.