Table 2.

Summary of Results and Interpretation in Feygina et al. (2010, Study 3) and Our close replication Based on Only the Original Study Conditions (Original Study Control vs. System-Preservation Condition).

	FEYGINA ET AL. (2010, STUDY 3)		OUR CLOSE REPLICATION
Pro-environmental intentions	Interpretation	Results reported	Interpretation	Results
Main effect of system justification	‘marginal’ … ‘in the expected direction’	b = 0.46, SE = 0.26, β = .36, t(37) = 1.75, p = .09.	No	b = 0.05, SE = 0.09, β = .04, t(374) = 0.59, p = .553
Main effect of condition	No	b = 0.20, SE = 0.52, β = .06, ns	No	b = –0.40, SE = 0.50, β = –.11, t(374) = –0.81, p = .421
Interaction between system justification and condition	Yes	b = 0.95, SE = 0.41, β = .49, t(37) = 2.29, p = .03	No	b = 0.05, SE = 0.12, β = .06, t(374) = 0.43, p = .670
Follow-up test 1 – comparing high (M + 1SD) versus low (M – 1SD) system justifiers
Difference between conditions for high system justifiers?	Yes; PEB intentions were more pro-environmental in the system-preservation condition than the control	b = 1.40, SE = 0.82, β = .44, t(37) = 1.72, p = .09	N/A – follow up tests not performed given non-significant interaction
Difference between conditions for low system justifiers?	No; PEB intentions were not reliably different across conditions.	b = –1.00, SE = 0.65, β = –.32, t(37) = –1.53, p > .10
Follow-up test 2 – examining relationship between system justification and PEB intentions in each condition
Control condition	‘A negative (albeit marginal) relationship’	b = –0.46, SE = 0.26, β = –.36, t(37) = 1.75, p < .09	N/A – follow up tests not performed given non-significant interaction
Experimental condition	‘negative relationship was eliminated’ … ‘no reliable relationship’	b = 0.49, SE = 0.32, β = .39, t(37) = 1.53, p = .14
Petitions	Interpretation	Results reported	Interpretation	Results
Main effect of system justification	‘marginal’, indicating system justification was inversely related to signing petitions	b = –0.64, SE = 0.38, Wald = 2.84, p = .09	Yes	b = –.35, SE = 0.10, 95% CI [–0.54, –0.16], Odds ratio (OR) = 0.70, Wald = 13.08, p < .001
Main effect of condition	No	b = 0.58, SE = 0.65, Wald = .82, ns	Yes	b = –1.36, SE = 0.54, 95% CI [–2.43, –0.31], OR = 0.26, Wald = 6.37, p = .012
Interaction between system justification and condition	Yes	b = 1.26, SE = 0.57, Wald = 5.01, p = .03	Yes	b = 0.37, SE = 0.13, 95% CI [0.12, 0.63], OR = 1.45, Wald = 8.20, p = .004
Follow-up test 1 – comparing high (M + 1SD) versus low (M – 1SD) system justifiers
Difference between conditions for high system justifiers?	No	None; inferred from statement ‘reframing environmentalism as supporting (rather than undermining) the American way or life eliminates the negative effect of system justification on pro-environmental behavior’. Graph shows cross-over interaction.	Yes. But note that robustness checks described in text indicate no significant difference at M + 1SD system justification; differences emerge at extremely high levels of system justification in exploratory analyses.	49.33% of high system justifiers (M + 1SD) sign most petitions in the system-preservation condition versus 34.40% in the control condition.
Difference between conditions for low system justifiers?	Yes		Possible backfire? But note that robustness checks indicate no significant difference at M – 1SD, or any low level of system justification.	60.59% of low system justifiers (M – 1SD) sign most petitions in the original study control condition versus 47.48% in the system-preservation condition.
Follow-up test 2 – examining relationship between system justification and petitions signed in each condition
Control condition	‘marginal negative effect of system justification on the probability of signing petitions’	None	Negative effect of system justification on the likelihood of signing petitions	b = –0.35, SE = 0.10, 95% Asymptotic CI [–0.54, –0.16], Wald = 13.08, p < .001
Experimental condition	‘no longer a reliable relationship’	b = 0.63, SE = 0.41, Wald = 2.33, ns	No effect	b = 0.02, SE = 0.09, 95% Asymptotic CI [–0.15, 0.20], Wald = 0.08, p = .783