https://orcid.org/0000-0001-6171-1381
Figures
Abstract
To investigate impression formation, researchers tend to rely on statements that describe a person’s behavior (e.g., “Alex ridicules people behind their backs”). These statements are presented to participants who then rate their impressions of the person. However, a corpus of behavior statements is costly to generate, and pre-existing corpora may be outdated and might not measure the dimension(s) of interest. The present study makes available a normed corpus of 160 contemporary behavior statements that were rated on 4 dimensions relevant to impression formation: morality, competence, informativeness, and believability. In addition, we show that the different dimensions are non-independent, exhibiting a range of linear and non-linear relationships, which may present a problem for past research. However, researchers interested in impression formation can control for these relationships (e.g., statistically) using the present corpus of behavior statements.
Citation: Mickelberg A, Walker B, Ecker UKH, Howe P, Perfors A, Fay N (2022) Impression formation stimuli: A corpus of behavior statements rated on morality, competence, informativeness, and believability. PLoS ONE 17(6): e0269393. https://doi.org/10.1371/journal.pone.0269393
Editor: Jeff Galak, Carnegie Mellon Univeristy, UNITED STATES
Received: March 30, 2022; Accepted: May 19, 2022; Published: June 3, 2022
Copyright: © 2022 Mickelberg et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data and materials are available on the Open Science Framework: https://osf.io/qnv95/.
Funding: The research was supported by a PhD scholarship funded by the Defence Science and Technology Group of the Department of Defence (A.M.) and a Defence Science and Technology (DST) Group Collaborative Agreement 8382 (N.F., B.W., A.P. and P.H.). The funders had no role in the study design, data collection and analysis or decision to publish. They did provide feedback on draft of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Without direct access to the inner thoughts and feelings of others, we often rely on behavioral information to form impressions of people. Some behaviors may elicit a positive impression (e.g., saving a drowning friend), whereas others may elicit a negative impression (e.g., having an extramarital affair). Each behavior serves as a building block in the impression formation process, and the impressions we form guide our social interactions with friends, colleagues, romantic partners, and casual acquaintances [e.g., 1, 2].
To investigate impression formation under controlled laboratory conditions, researchers often present participants with statements that describe a person’s behavior (e.g., “Alex ridicules people behind their backs”), and then participants rate their impressions of that person [e.g., 3–5]. Researchers have primarily focused on dimensions that capture important facets of a person’s character: (i) morality [also called communion, see 6], encompassing honesty, loyalty, and cooperativeness, and (ii) competence [also called agency, see 6], encompassing intelligence, efficiency, and capability [7, 8]. It should be noted that some researchers use the label ‘warmth’ interchangeably with morality [e.g., 7, 9–11] while others argue that warmth is an overarching factor encompassing morality and sociability [12, 13, see also 6]. We follow Brambilla et al. [14, 15] with morality being core to impression formation. Moral behaviors (e.g., “she kept a friend’s secret”, “he lied to his parents”) indicate whether a person’s intentions are good or bad, while competence behaviors (e.g., “they achieved a challenging goal”, “she did not get good marks at university”) indicate their ability to successfully execute a task [11, 16]. Although both dimensions guide impression formation, moral behaviors are found to be more influential than competence behaviors [13, 14, 17, see also 15]. While morality and competence are the major dimensions investigated to date, other dimensions also play a role.
Another dimension that guides impression formation is informativeness. Behavior statements that are high in informativeness are diagnostic of a person’s true character, resulting in greater impression change [18, 19]. Research has shown that the informativeness dimension is related to other dimensions: behavior statements that are morally negative are rated as more informative than morally positive statements [e.g., 20, 21] and morally extreme behavior statements are rated as more informative than morally moderate statements [22–24, see 25 for a review]. It has recently been established that the believability of behavioral information is also important to impression formation; person impressions are updated only when the information is considered to be believable, regardless of how informative or extreme the information is [5]. Thus, believability may moderate the effect of the other dimensions known to guide impression formation [see also 26].
To examine how these dimensions inform person impressions, researchers require a corpus of behavior statements that vary on the relevant dimensions [27–30]. To avoid the cost associated with generating a corpus of statements, it is common to use behavior statements that were generated in prior studies [e.g., 5, 17, 19, 31, 32]. However, doing so can be problematic. First, if the behavior statements were rated by a small sample of judges, they may measure the dimensions of interest imprecisely. Second, behavior statements can become outdated, which can make them difficult for participants to evaluate [e.g., whether “replaced the ribbon on his typewriter” indicates competence; see 28] and may limit their contemporary real-world applicability [e.g., whether someone “had difficulty balancing a checkbook” is unlikely to come up in the present day; see 28]. Third, researchers may be interested in dimensions that were not assessed in past studies—for instance, the statements generated by Chadwick et al. [27] and Fuhrman et al. [28] were not rated on informativeness or believability.
To address these issues, we generated a comprehensive and contemporary list of 160 behavior statements that were rated by a large sample of judges (N = 400). The statements were rated on four dimensions: morality, competence, informativeness, and believability. In the present study, the behavior statements were designed to vary across the morality dimension (from extreme positive, e.g., “Person X sold their house to fund a local program for the needy”, to extreme negative, e.g., “Person X kicked their pet dog hard in the head because it didn’t come when called”) and the competence dimension (from extreme positive, e.g., “Person X did all the repair work on their car”, to extreme negative, e.g., “Person X failed their driver’s license test for the fourth time”). This included statements that were designed to be neutral on both dimensions (e.g., “Person X buys a loaf of bread every day, as they love the smell of freshly baked bread in the morning”). We anticipated that the behavior statements would naturally vary on the informativeness and believability dimensions.
We first present the statements and their ratings across the four dimensions of interest. We then examine the relationships between the four dimensions. Any relationships would highlight potential confounds that should be taken into account by researchers. The corpus provides a normed set of contemporary behavior statements that enables researchers to test new research questions in impression formation.
Method
The study was conducted in accordance with the National Statement on Ethical Conduct in Human Research [33]. It was approved by the University of Western Australia’s Human Research Ethics Office. Participants viewed an approved information sheet before giving informed consent to take part.
Participants
A convenience sample of participants were recruited from the United States via the online crowd-sourcing platform Prolific. The sample comprised N = 400 participants (female: 205; male: 189; other: 5; prefer not to say: 1) with an age range of 18–73 years (M = 33.66, SD = 11.66). Each participant received the equivalent of £1.50 (approximately US$2) upon completion of the study.
Behavior statements
The study used a pool of N = 160 behavior statements. These included behaviors generated by the authors (n = 94), with the remainder (n = 66) adapted from prior studies (14,26,27,33). The behavior statements took the form of “Person X…”, describing a behavior in which Person X is the agent. The behaviors were designed to vary in morality (positive, negative) and competence (positive, negative), including behaviors that tended toward neutral on both dimensions. Moral behaviors were generated with reference to three of the five psychological foundations of morality: harm/care, fairness/reciprocity, and ingroup/loyalty [34, note however that other conceptualisations of morality also exist, e.g., 35, 36].
To help ensure sufficient variation across each dimension, the authors brainstormed statements from five categories: positive morality (48 statements; e.g., “Person X sold their house to fund a local program for the needy”), negative morality (48 statements; e.g., “Person X set fire to the community hall in the middle of the night”), positive competence (20 statements; e.g., “Person X solved a crossword puzzle in the newspaper”), negative competence (20 statements; e.g., “Person X forgot to turn off the bathwater, flooding the house”), and neutral (24 statements; e.g., “Person X went to a friend’s house to play a card game”). Statements in the positive and negative competence categories were designed to be neutral on the morality dimension. Although not intentionally designed to vary on informativeness or believability, it was anticipated that the behavior statements would vary on these dimensions.
Each participant was presented with 40 statements selected randomly subject to the following constraints: 12 from each of the positive and negative morality categories, 5 from each of the positive and negative competence categories, and 6 from the neutral category. Pre-testing indicated that participants could become fatigued if they rated more than 40 statements. Behavior statements were sampled such that each statement was rated by 100 participants.
Procedure
The study was performed online using an internet-enabled device, and took approximately 15 minutes to complete. After participants provided informed consent, they supplied their age and gender, and read over the instructions. The instructions explained that they would rate 40 behavior statements on various dimensions, with each statement describing a different person (e.g., Person 1, Person 2; this reduced the possibility of statements interacting with each other). The behavior statements were then presented in a random order. Participants rated each statement on its morality (“How morally bad or good is the behavior described in the statement?”), from -4 (very morally bad) to 4 (very morally good), with 0 indicating neutral; its competence (“How would you rate the person’s competence from the behavior described in the statement?”), from -4 (very incompetent) to 4 (very competent), with 0 indicating neutral; its informativeness (“How informative is the statement? How valuable is it when forming an impression of the person?”), from 0 (not informative) to 8 (very informative); and its believability (“How believable is the statement? To what extent could it happen in real life?”), from 0 (not believable) to 8 (very believable). Ratings were entered using horizontally aligned radio buttons. Participants were then debriefed.
Results
All analyses were performed and all figures created in R [37]. Data visualizations were created using ggplot2 [38], the raincloud plot package [39], and corrplot [40]. The data and R Script are available on the Open Science Framework: https://osf.io/qnv95/.
Preliminary analysis
To identify uniform responding, we calculated each participant’s standard deviation across all measures (morality, competence, informativeness, and believability). No outliers were identified using the interquartile rule with a 2.2 multiplier (i.e., cutoff = SD < Q1–2.2 × IQR) [see 41]. In addition, each measure was approximately normally distributed (|skew| < 2 and |kurtosis| < 9).
Behavior statement ratings
The distributions of the mean morality, competence, informativeness, and believability ratings for each behavior statement are shown in Fig 1. The ratings ranged across the entire morality dimension (Fig 1A) and most of the competence dimension (Fig 1B). Informativeness ratings varied substantially (Fig 1C). Believability ratings varied considerably (Fig 1D), but most statements were rated as believable (i.e., in the upper part of the scale). Ratings for the full corpus of behavior statements are given in Table 1, and an interactive version of Table 1 is available on the Open Science Framework: https://osf.io/jv7fk.
Mean morality (A), competence (B), informativeness (C), and believability (D) ratings for each behavior statement. The “cloud” shows the density distribution for the given ratings. Each dot point shows the mean rating for a single behavior statement; points are jittered vertically to avoid overplotting. Boxplots show the first to third quartiles, the bolded vertical line denotes the median, and the whiskers denote 1.5 times the interquartile range.
Relationships between the dimensions
There were moderate-to-strong Pearson correlations between the morality and competence ratings, the morality and believability ratings, the competence and believability ratings, and the informativeness and believability ratings (see Fig 2).
The color shows the direction of the relationship, with positive in blue and negative in orange. Circle size shows the strength of the relationship, with a larger circle indicating a stronger relationship. Note *p < .050, **p < .010, ***p < .001.
The relationships between each of the dimensions are visualized in Fig 3. Inspection of the figure indicated linear and non-linear relationships between several pairs of dimensions. We therefore tested for linear and quadratic relationships using orthogonal polynomial regression (see Table 2 for statistical output). The morality and competence ratings showed a strong positive linear relationship, indicating that behavior statements rated as more positive in morality were rated as more competent (see Fig 3A). The morality and informativeness ratings showed a strong quadratic effect, indicating that behavior statements rated as more extreme in morality (negative or positive) were rated as more informative (see Fig 3B). The morality and believability ratings showed both a moderate positive linear relationship and a quadratic relationship (see Fig 3C). The linear effect indicates that behavior statements rated as more positive in morality were rated as more believable, while the quadratic effect indicates that behavior statements rated as more extreme in morality (negative or positive) were rated as less believable.
Scatter plots depicting the association between (A) morality and competence, (B) morality and informativeness, (C) morality and believability, (D) competence and informativeness, (E) competence and believability, and (F) informativeness and believability. Dot points represent the mean ratings for each behavior statement. The green lines show the linear trends and the blue lines show the quadratic trends. The shaded areas show the 95% confidence intervals. Rugs (i.e., the blue lines along the x- and y-axes) show distribution density.
The competence and informativeness ratings showed a strong quadratic effect (see Fig 3D), indicating that behavior statements rated as more extreme in competence (negative or positive) were rated as more informative. The competence and believability ratings showed a moderate positive linear relationship and a quadratic relationship (see Fig 3E). The linear effect indicates that behavior statements rated as more positive in competence were rated as more believable, while the quadratic effect indicates that behavior statements rated as more extreme in competence (negative or positive) were rated as less believable. The informativeness and believability ratings showed a strong negative linear relationship, indicating that behavior statements rated as more informative were rated as less believable (see Fig 3F).
Discussion
The present study provides a normed corpus of 160 contemporary behavior statements. Each behavior statement was rated on the dimensions of morality [11], competence [10], informativeness [23], and believability [5], which are known to affect impression formation. The behavior statement ratings varied widely on the morality, competence, and informativeness dimensions, providing researchers with substantial scope to investigate the effects of these dimensions on impression formation. There was less variation on the believability dimension, with most behavior statements rated as being at least moderately believable. Given that behavior statements need to be believable to affect person impressions [5], the general believability of the behavior statements should be advantageous to researchers using the corpus.
Researchers interested in the influence of specific dimensions on impression formation may need to control for the contribution of related dimensions. Our results indicate a range of linear and quadratic relationships between the morality, competence, informativeness, and believability dimensions. The morality and competence dimensions showed a positive linear relationship, indicating that more morally positive behavior statements were rated as more competent. This replicates prior research [29, 42, 43], and suggests a halo effect [10, 44] whereby favorable judgments on the morality dimension positively influence judgements on the competence dimension (or vice versa). The informativeness dimension showed a quadratic relationship to the morality and competence dimensions: behavior statements rated as more extreme in morality or competence (i.e., extreme positive or extreme negative) were associated with an increase in informativeness. These findings are consistent with an extremity bias, whereby more morally extreme information is given greater weight in impression formation [24, 45, 46].
The believability dimension showed positive linear relationships and quadratic relationships with the morality and competence dimensions. Behavior statements rated as more positive in morality/competence were generally rated as more believable (than more negative statements), and more extreme (positive/negative) behaviors were associated with a decrease in believability. These relationships may be explained by people’s expectations, in so far as people expect others to behave in positive and non-extreme ways [e.g., person positivity bias, see 47] so find such behaviors more believable. Our final test showed a strong negative linear relationship between informativeness and believability, indicating that more informative behavior statements were also rated as less believable. Together, these findings make intuitive sense, suggesting that more unexpected and surprising behaviors, which are less believable, are considered to be more informative [19, see also “frequency-weight” theories, 22]. The negative relationship between informativeness and believability is also consistent with recent research on misinformation (e.g., fake news and conspiracy theories). Even if low in believability, misinformation can be perceived to be ‘informative if true’, and therefore has the potential to strongly sway opinion [48, 49] and be widely shared online [50, see also 51].
To conclude, the present study provides a normed corpus of 160 contemporary behavior statements. The statements were rated by a large sample of judges (N = 400, with each behavior statement rated by 100 judges) on four dimensions relevant to impression formation: morality, competence, informativeness, and believability. Importantly, the different dimensions were non-independent; a range of linear and non-linear relationships between the dimensions were identified. Accounting for these relationships (e.g., statistically) can help researchers avoid drawing unwarranted conclusions. For example, researchers investigating the effect of competence on impression formation may find their results are better explained by morality [e.g., see 52] or that the effect of a specific dimension is moderated by statement informativeness or believability. Given these considerations, we believe the corpus of behavior statements generated in the present study will be valuable to researchers interested in impression formation.
Acknowledgments
We are grateful to Martin Wood and David Kernot (Defence Science and Technology Group) for their valuable feedback on early drafts of the manuscript.
References
- 1. Asch SE. Forming impressions of personality. J Abnorm Soc Psychol. 1946;258–90. pmid:20995551
- 2. Srull TK, Wyer RS. Person memory and judgment. Psychol Rev. 1989;96(1):58–83. pmid:2648446
- 3. Brannon SM, Sacchi DLM, Gawronski B. (In)consistency in the eye of the beholder: The roles of warmth, competence, and valence in lay perceptions of inconsistency. J Exp Soc Psychol. 2017;70:80–94.
- 4. Rydell RJ, McConnell AR, Strain LM, Claypool HM, Hugenberg K. Implicit and explicit attitudes respond differently to increasing amounts of counterattitudinal information. Eur J Soc Psychol. 2007;37(5):867–78.
- 5. Cone J, Flaharty K, Ferguson MJ. Believability of evidence matters for correcting social impressions. PNAS Proc Natl Acad Sci U S Am. 2019;116(20):9802–7. pmid:31036663
- 6. Abele AE, Wojciszke B. Agency and communion from the perspective of self versus others. J Pers Soc Psychol. 2007;93(5):751–63. pmid:17983298
- 7. Fiske ST, Cuddy AJC, Glick P, Xu J. A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. J Pers Soc Psychol. 2002;82(6):878–902. pmid:12051578
- 8. Wojciszke B. Multiple meanings of behavior: Construing actions in terms of competence or morality. J Pers Soc Psychol. 1994;67(2):222–32.
- 9. Fiske ST. Stereotype content: Warmth and competence endure. Curr Dir Psychol Sci. 2018;27(2):67–73. pmid:29755213
- 10. Fiske ST, Cuddy AJC, Glick P. Universal dimensions of social cognition: Warmth and competence. Trends Cogn Sci. 2007;11(2):77–83. pmid:17188552
- 11. Morality Wojciszke B. and competence in person- and self-perception. Eur Rev Soc Psychol. 2005;16:155–88.
- 12. Brambilla M, Leach CW. On the importance of being moral: The distinctive role of morality in social judgment. Soc Cogn. 2014;32(4):397–408.
- 13. Goodwin GP, Piazza J, Rozin P. Moral character predominates in person perception and evaluation. J Pers Soc Psychol. 2014;106(1):148–68. pmid:24274087
- 14. Brambilla M, Carraro L, Castelli L, Sacchi S. Changing impressions: Moral character dominates impression updating. J Exp Soc Psychol. 2019;82:64–73.
- 15.
Brambilla M, Sacchi S, Rusconi P, Goodwin GP. Chapter Four—The primacy of morality in impression development: Theory, research, and future directions. In: Gawronski B, editor. Advances in Experimental Social Psychology. Academic Press; 2021. p. 187–262.
- 16. De Bruin ENM, Van Lange PAM. What people look for in others: Influences of the perceiver and the perceived on information selection. Pers Soc Psychol Bull. 2000;26(2):206–19.
- 17. Ybarra O. When first impressions don’t last: The role of isolation and adaptation processes in the revision of evaluative impressions. Soc Cogn. 2001;19(5):491–520.
- 18. Cone J, Ferguson MJ. He did what? The role of diagnosticity in revising implicit evaluations. J Pers Soc Psychol. 2015;108(1):37–57. pmid:25365037
- 19. Mende-Siedlecki P, Baron S, Todorov A. Diagnostic value underlies asymmetric updating of impressions in the morality and ability domains. J Neurosci. 2013;33(50):19406–15. pmid:24336707
- 20. Peeters G, Czapiński J. Positive-negative asymmetry in evaluations: The distinction between affective and informational negativity effects. In: European Review of Social Psychology. 1990. p. 33–60.
- 21. Lupfer M, Dupuis S. How pervasive is the negativity bias in judgments based on character appraisal? Pers Soc Psychol Bull. 2000;26:1353–66.
- 22. Fiske ST. Attention and weight in person perception: The impact of negative and extreme behavior. J Pers Soc Psychol. 1980;38:889–906.
- 23. Skowronski JJ, Carlston DE. Negativity and extremity biases in impression formation: A review of explanations. Psychol Bull. 1989;105(1):131–42.
- 24. Skowronski JJ, Carlston DE. Social judgment and social memory: The role of cue diagnosticity in negativity, positivity, and extremity biases. J Pers Soc Psychol. 1987;52(4):689–99.
- 25. Rusconi P, Sacchi S, Brambilla M, Capellini R, Cherubini P. Being honest and acting consistently: Boundary conditions of the negativity effect in the attribution of morality. Soc Cogn. 2020;38(2):146–78.
- 26. Guillory JJ, Geraci L. Correcting erroneous inferences in memory: The role of source credibility. J Appl Res Mem Cogn. 2013;2(4):201–9.
- 27. Chadwick RA, Bromgard G, Bromgard I, Trafimow D. An index of specific behaviors in the moral domain. Behav Res Methods. 2006;38(4):692–7. pmid:17393841
- 28. Fuhrman RW, Bodenhausen GV, Lichtenstein M. On the trait implications of social behaviors: Kindness, intelligence, goodness, and normality ratings for 400 behavior statements. Behav Res Methods Instrum Comput. 1989;21(6):587–97.
- 29. Rosenberg S, Nelson C, Vivekananthan PS. A multidimensional approach to the structure of personality impressions. J Pers. 1968 Aug;9(4):283–94.
- 30. Piazza J, Sousa P, Rottman J, Syropoulos S. Which appraisals are foundational to moral judgment? Harm, injustice, and beyond. Soc Psychol Personal Sci. 2019;10(7):903–13.
- 31. Reeder GD, Coovert MD. Revising an impression of morality. Soc Cogn. 1986;4(1):1–17.
- 32. Vonk R. Effects of inconsistent behaviors on person impressions: A multidimensional study. Pers Soc Psychol Bull. 1995;21(7):674–85.
- 33. NHMRC. National statement on ethical conduct in human research (2007) [Internet]. 2018 [cited 2022 Mar 29]. Available from: https://www.nhmrc.gov.au/about-us/publications/national-statement-ethical-conduct-human-research-2007-updated-2018#toc__725
- 34. Graham J, Nosek BA, Haidt J, Iyer R, Koleva S, Ditto PH. Mapping the moral domain. J Pers Soc Psychol. 2011;101(2):366–85. pmid:21244182
- 35. Janoff-Bulman R, Carnes N. Surveying the moral landscape. Personal Soc Psychol Rev Off J Soc Personal Soc Psychol Inc. 2013;17(3):219–36. pmid:23504824
- 36. Rai TS, Fiske AP. Moral psychology is relationship regulation: Moral motives for unity, hierarchy, equality, and proportionality. Psychol Rev. 2011;118(1):57–75. pmid:21244187
- 37.
R Core Team. R: A language and environment for statistical computing [Internet]. R Foundation for Statistical Computing. 2021. Available from: https://www.r-project.org/
- 38.
Wickham H. ggplot2: Elegant graphics for data analysis [Internet]. Springer-Verlag New York; 2016. Available from: https://ggplot2-book.org/
- 39. Allen M, Poggiali D, Whitaker K, Marshall TR, Kievit RA. Raincloud plots: A multi-platform tool for robust data visualization. Wellcome Open Res. 2019;4:63. pmid:31069261
- 40. Wei T, Simko V. Package “corrplot”: Visualization of a correlation matrix [Internet]. 2016. Available from: https://github.com/taiyun/corrplot
- 41. Hoaglin D, Iglewicz B. Fine-tuning some resistant rules for outlier labeling. J Am Stat Assoc. 1987;82(400):1147–9.
- 42. Judd CM, James-Hawkins L, Yzerbyt V, Kashima Y. Fundamental dimensions of social judgment: Understanding the relations between judgments of competence and warmth. J Pers Soc Psychol. 2005;89(6):899–913. pmid:16393023
- 43. Kervyn N, Yzerbyt V, Judd CM. Compensation between warmth and competence: Antecedents and consequences of a negative relation between the two fundamental dimensions of social perception. Eur Rev Soc Psychol. 2010;21(1):155–87.
- 44. Thorndike EL. A constant error in psychological ratings. J Appl Psychol. 1920;4(1):25–9.
- 45. Reeder GD, Brewer MB. A schematic model of dispositional attribution in interpersonal perception. Psychol Rev. 1979;86(1):61–79.
- 46. Rothbart M, Park B. On the confirmability and disconfirmability of trait concepts. J Pers Soc Psychol. 1986;50(1):131–42.
- 47. Sears DO. The person-positivity bias. J Pers Soc Psychol. 1983;44(2):233–50.
- 48. Swire‐Thompson B, Ecker UKH, Lewandowsky S, Berinsky A. They might be a liar but they’re my liar: Source evaluation and the prevalence of misinformation. Polit Psychol. 2019;41(1):21–34.
- 49. Thorson EA. Belief echoes: The persistent effects of corrected misinformation. 2016;33(3):460–80.
- 50. Vosoughi S, Roy D, Aral S. The spread of true and false news online. Science. 2018;359(6380):1146–51. pmid:29590045
- 51. Altay S, Araujo E de, Mercier H. “If this account is true, it is most enormously wonderful”: Interestingness-if-true and the sharing of true and false news. Digit Journal. 2021;10(2):1–22.
- 52. Stellar JE, Willer R. Unethical and inept? The influence of moral information on perceptions of competence. J Pers Soc Psychol. 2018;114(2):195–210. pmid:29389214