Table 3.

Potential moderators we explored (and failed to detect) that we hypothesized could moderate the effect of having people sign a veracity statement attesting to their honest reporting placed before versus after reporting

Potential moderator	Relevant studies
Reporting online via typing vs. reporting on paper via handwriting	Electronic reporting vs. handwriting reporting (studies 1 and 2)
Study population	Laboratory population (Boston, studies 1,2, and 6; and Chicago, study 6; both community and student populations in Boston and Chicago) vs. MTurk (studies 3–5)
Verbal vs. written instructions	Study instructions written on computer screen/paper and participants read on their own (studies 1 and studies 3–5) vs. research assistant provided instructions out loud (studies 2 and 6)
Task type	Die rolling (study 1), anagram/word scramble (studies 2–5), matrix task (study 6), and expense reporting (study 6)
Incentive/amount of additional money at stake ($50 and under)	Raffle for $50 (study 1), up to $10 (study 2), $0.10 per reported answer (study 3 and 4), $0.30 per reported answer (5), and up to $42 for reported answer and reported expenses
Amount of baseline cheating	Study 3 (lowest cheating rate, 23% in the control group) to study 6 (highest cheating rate, 56% in the control group)
Type of reporting form	Regular participation form where generally there is no expectation of an honesty prompt (studies 1–5) vs. official-looking tax form where in naturalistic context there is a general expectation of an honesty prompt (study 6)