When misinformation is corrected, the correction often fails to fully remove the misinformation's influence on memory and reasoning — the continued influence effect. A contested further claim is that corrections can be differentially ineffective, or even backfire (increase belief in the corrected claim), when the misinformation is congruent with the recipient's political worldview, with an asymmetry reported for conservative partisans. Ecker and Ang (2019) reported exactly this pattern in an Australian continued-influence paradigm using a single misconduct topic (political embezzlement): retractions were entirely ineffective for conservative participants when worldview-congruent misinformation was corrected, with at most a marginal, statistically unreliable backfire trend. The open problem asks: under what conditions and through what mechanisms do corrections of worldview-congruent misinformation differentially fail or backfire for partisans, and do such effects reflect genuine ironic belief change or strategic/expressive responding? Subsequent large-scale experiments, direct replications, and meta-analyses have mostly failed to find worldview backfire, while observational work continues to find ideological asymmetry after corrections, so the boundary conditions, mechanisms, and psychological locus of worldview-motivated correction resistance remain open.

Background

Misinformation is false or inaccurate information circulated as fact, and a correction (also called a retraction or debunking) is a message that explicitly negates an earlier false claim. A robust experimental finding is the continued influence effect (CIE): even after people receive and remember a correction, the corrected misinformation keeps shaping their memory and inferential reasoning (Johnson & Seifert, 1994). A controversial question is whether corrections work equally well for everyone. Political psychology predicts they may not: motivated reasoning — the tendency to evaluate evidence so as to protect prior attitudes — suggests that when a correction threatens a cherished political attitude, it may be discounted, ignored, or even counterproductively strengthen the false belief. The latter outcome is the worldview backfire effect: a correction that increases belief in the very misconception it targets. It was first reported by Nyhan and Reifler (2010), whose corrections of claims about weapons of mass destruction in Iraq appeared to increase misperceptions among conservative participants, and was publicized in the influential review by Lewandowsky, Ecker, Seifert, Schwarz, and Cook (2012). Many later studies, however, found little or no effect of attitudes or partisanship on how corrections are processed (e.g., Aird et al., 2018; Gordon, Ecker, & Lewandowsky, 2019), producing an unresolved controversy over whether worldview backfire exists at all.

Ecker and Ang (2019) built the paradigm at the center of this problem to test the question directly. They presented self-identified partisans of the two major Australian parties (left-leaning Labor and right-leaning Liberal) with fictional news reports alleging misconduct — specifically, misappropriation of public funds — by politicians of one or the other party, in a fully crossed design, and then either retracted the allegation or let it stand. Two experiments distinguished a singular-event claim (a named MP misappropriated funds) from a general assertion (Labor politicians are three times as likely as Liberals to misappropriate funds). The key measures were references to the misinformation in open-ended inferential reasoning (e.g., who should be investigated, what conclusions follow) and rating-scale belief questions. They found that retractions were less effective when the misinformation was congruent with participants' political worldview, that this resistance was pronounced particularly for conservative participants, and that in the general-assertion, worldview-congruent condition for conservatives the retraction was entirely ineffective, with at most a marginal, statistically unreliable backfire trend rather than a significant backfire effect. The study used one misconduct topic (embezzlement), one national context, and an undergraduate sample. Whether the asymmetry generalizes to other materials, domains, and populations, what mechanism produces it (candidate factors include the moral acceptability of the misconduct, the salience of the misinformation, the strength of party identification, and the centrality of the political attitude to the person's identity — the last explicitly proposed as a mediator in the authors' own follow-up replication, Ecker, Sze, & Andreotta, 2021), whether genuine worldview backfire exists and under what boundary conditions, and whether the observed resistance reflects genuine ironic belief change or strategic expression rather than true belief (e.g., adjusted belief expressions that deviate from true beliefs under social-norm pressures) all remain open.

Problem Statement

The problem is to determine when, for whom, and why corrections of politically congruent misinformation fail or backfire, and what such resistance psychologically is. Concretely, characterizing worldview-motivated resistance to misinformation corrections raises three linked questions, which are jointly required for a full resolution (a rigorous answer to any one of them alone is a partial contribution and leaves the others open): (1) Boundary conditions and generalization: under what conditions (type of misconduct and its moral acceptability, salience of the misinformation, singular-event versus general-assertion claims, real-world versus fictional materials, acute identity threat), and in materials and populations beyond the original single embezzlement topic and Australian undergraduate sample, are corrections of worldview-congruent misinformation differentially ineffective for partisans, and do genuine ironic belief increases (worldview backfire) ever occur, given that large-scale experiments and direct replications have so far mostly found none? A well-supported null — meaning one that rests on more than one adequately powered study rather than on a single experiment, and that shows no differential correction failure and no genuine backfire across a stated, bounded set of conditions — answers this question just as a positive finding of differential resistance or backfire would, provided the bound is stated explicitly; conditions left outside that bound remain open. (2) Mechanisms: which factors produce the asymmetry between conservative and liberal partisans when it is observed? Candidate moderators and mediators to be tested include the moral acceptability of the misconduct, the salience of the misinformation, the strength of party identification, and the centrality of the political attitude to the person's identity (the last explicitly proposed as a mediator in the direct replication, Ecker, Sze, & Andreotta, 2021); a well-supported definitive finding that none of the tested candidate mechanisms produces an observed asymmetry answers this question just as a positive mechanism finding would, provided each candidate mechanism was actually tested under conditions able to reveal the asymmetry; the mere absence of an asymmetry, without such mechanism-level testing, does not answer this sub-question. (3) Belief versus expression: do differential resistance and any backfire reflect genuine belief updating, or strategic/expressive responding — adjusted belief expressions that deviate from true beliefs under social-norm pressures? A well-supported finding that resolves this distinction either way — that the resistance reflects genuine belief updating, or that it reflects strategic/expressive responding — answers this question in full. Consistent with the source line's own framing, no single experiment — even an adequately powered direct replication — settles the existence-and-boundary-conditions question: the direct replication's authors keep open that worldview backfire may still occur under specific conditions (e.g., less contrived real-world misinformation, retractions perceived as acutely identity-threatening).

References

  1. Ecker, U. K. H., & Ang, L. C. (2019). Political attitudes and the processing of misinformation corrections. Political Psychology, 40(2), 241–260. https://doi.org/10.1111/pops.12494
  2. Johnson, H. M., & Seifert, C. M. (1994). Sources of the continued influence effect: When misinformation in memory affects later inferences. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20(6), 1420–1436. https://doi.org/10.1037/0278-7393.20.6.1420
  3. Nyhan, B., & Reifler, J. (2010). When corrections fail: The persistence of political misperceptions. Political Behavior, 32(2), 303–330. https://doi.org/10.1007/s11109-010-9112-2
  4. Lewandowsky, S., Ecker, U. K. H., Seifert, C. M., Schwarz, N., & Cook, J. (2012). Misinformation and its correction: Continued influence and successful debiasing. Psychological Science in the Public Interest, 13(3), 106–131. https://doi.org/10.1177/1529100612451018
  5. Aird, M. J., Ecker, U. K. H., Swire, B., Berinsky, A. J., & Lewandowsky, S. (2018). Does truth matter to voters? The effects of correcting political misinformation in an Australian sample. Royal Society Open Science, 5(12), 180593. https://doi.org/10.1098/rsos.180593
  6. Wood, T., & Porter, E. (2019). The elusive backfire effect: Mass attitudes' steadfast factual adherence. Political Behavior, 41(1), 135–163. https://doi.org/10.1007/s11109-018-9443-y
  7. Gordon, A., Ecker, U. K. H., & Lewandowsky, S. (2019). Polarity and attitude effects in the continued-influence paradigm. Journal of Memory and Language, 108, 104028. https://doi.org/10.1016/j.jml.2019.104028
  8. Swire-Thompson, B., DeGutis, J., & Lazer, D. (2020). Searching for the backfire effect: Measurement and design considerations. Journal of Applied Research in Memory and Cognition, 9(3), 286–299. https://doi.org/10.1016/j.jarmac.2020.06.006
  9. Ecker, U. K. H., Sze, B. K. N., & Andreotta, M. (2021). Corrections of political misinformation: No evidence for an effect of partisan worldview in a US convenience sample. Philosophical Transactions of the Royal Society B, 376(1822), 20200145. https://doi.org/10.1098/rstb.2020.0145
  10. Ecker, U. K. H., Lewandowsky, S., Cook, J., Schmid, P., Fazio, L. K., Brashier, N., Kendeou, P., Vraga, E. K., & Amazeen, M. A. (2022). The psychological drivers of misinformation belief and its resistance to correction. Nature Reviews Psychology, 1(1), 13–29. https://doi.org/10.1038/s44159-021-00006-y
  11. Swire-Thompson, B., Miklaucic, N., Wihbey, J. P., Lazer, D., & DeGutis, J. (2022). The backfire effect after correcting misinformation is strongly associated with reliability. Journal of Experimental Psychology: General, 151(7), 1655–1665. https://doi.org/10.1037/xge0001131
  12. Prike, T., Blackley, P., Swire-Thompson, B., & Ecker, U. K. H. (2023). Examining the replicability of backfire effects after standalone corrections. Cognitive Research: Principles and Implications, 8(1), 39. https://doi.org/10.1186/s41235-023-00492-z
  13. Prike, T., Reason, R., Ecker, U. K. H., Swire-Thompson, B., & Lewandowsky, S. (2023). Would I lie to you? Party affiliation is more important than Brexit in processing political misinformation. Royal Society Open Science, 10(2), 220508. https://doi.org/10.1098/rsos.220508
  14. Chan, M.-p. S., & Albarracín, D. (2023). A meta-analysis of correction effects in science-relevant misinformation. Nature Human Behaviour, 7(9), 1514–1525. https://doi.org/10.1038/s41562-023-01623-8
  15. DeVerna, M. R., Guess, A. M., Berinsky, A. J., Tucker, J. A., & Jost, J. T. (2024). Rumors in Retweet: Ideological asymmetry in the failure to correct misinformation. Personality and Social Psychology Bulletin, 50(1), 3–17. https://doi.org/10.1177/01461672221114222
  16. Butler, L. H., DeGutis, J., Tay, L. Q., Ecker, U. K. H., & Swire-Thompson, B. (2025). Corrections are effective for science misinformation. Nature Human Behaviour, 9(12), 2458–2460. https://doi.org/10.1038/s41562-025-02245-y
  17. Devaney, M., Federico, C. M., & Borgida, E. (2026). Debunking misinformation on critical race theory. Political Psychology, 47(2), e70065. https://doi.org/10.1111/pops.70065
  18. Quaranta, M., & Arrigoni, L. M. (2026). Politicians' misinformation, its correction, and partisanship in Italy. Italian Political Science Review / Rivista Italiana di Scienza Politica, 56(2), 112–125. https://doi.org/10.1017/ipo.2025.10084
  19. Laurent, V., Rusinek, S., & Kosinski, T. (2026). Continued influence effect: A three-dimensional framework shaping practical and theoretical perspectives. Psychological Reports. Advance online publication. https://doi.org/10.1177/00332941261464066