Armand ChatardPsychologie · Recherche · Société
Travaux en cours d’évaluation

Articles soumis

Cette rubrique présente mes travaux de recherche les plus récents, actuellement soumis à des revues scientifiques. Elle permet de suivre l’avancement de mes recherches avant leur publication officielle. Ces manuscrits étant encore en cours d’évaluation par les pairs, leurs résultats et leurs conclusions sont susceptibles d’évoluer.

2026 · Manuscrit soumis

Why Victims’ Emotions Fail to Inhibit Harm: The Exploitative Disinhibition Hypothesis

Chatard, A., Leblanc, P.-M., Delbreil, A., Harika-Germaneau, G., Jaafari, N., Xavier, J., Cailleau, V., Eymin, C., Cachau, P., Deliencourt, G., & Davignon, G. (2026). Why Victims’ Emotions Fail to Inhibit Harm: The Exploitative Disinhibition Hypothesis [Manuscript submitted for publication]. Université de Poitiers.

Lire l’abstract / Résumé

Why Victims’ Emotions Fail to Inhibit Harm: The Exploitative Disinhibition Hypothesis

Exploitative motivation toward a victim may bias the interpretation of that victim’s emotional signals, weakening their inhibitory force and facilitating harm. Two studies tested this exploitative disinhibition hypothesis in a context in which victims’ distress should strongly constrain behavior: sexual violence against children. In Study 1, a meta-analysis showed that child sexual offenders—whose behavior reflects a strong exploitative orientation toward children or adolescents—displayed poorer decoding of emotional signals than comparison groups (9 studies, 225 offenders). Study 2 tested the motivational–decoding link at the individual level in 63 incarcerated men convicted of sexual offenses against children or adolescents. Stronger implicit child–sex associations, used as an indirect marker of exploitative motivation, were linked to poorer emotion recognition, which in turn was linked to multiple victims. Together, these findings challenge a purely cognitive-deficit account of socio-affective decoding by showing that exploitative motivation may bias how victims’ emotions are interpreted.

Keywords: Emotion recognition · Exploitative motivation · Child sexual offending · Implicit associations · Disinhibition

2026 · Manuscrit soumis

Cruelty Without Dehumanization: Psychopathy, Sadism, and Victim Humanization

Chatard, A., Harika-Germaneau, G., Jaafari, N., & Delbreil, A. (2026). Cruelty Without Dehumanization: Psychopathy, Sadism, and Victim Humanization [Manuscript submitted for publication]. Université de Poitiers.

Lire l’abstract / Résumé

Cruelty Without Dehumanization: Psychopathy, Sadism, and Victim Humanization

Contrary to the dominant view, cruelty may not always require dehumanization. Because severe perpetrators often display high levels of psychopathy and sadism, they may derive greater satisfaction from inflicting suffering on fully human victims. We therefore predicted that men most prone to severe interpersonal violence would recognize victims’ humanity rather than deny it. This hypothesis was tested across four studies combining 344 forensic medico-legal records with community samples of 1,127 men in long-term heterosexual relationships. In archival records of intimate partner violence, dehumanizing insults peaked at intermediate severity but declined among perpetrators causing the most serious injuries. In US and UK samples, men scoring higher on both psychopathy and sadism reported greater physical partner violence and showed greater implicit humanization of injured women. Bayesian cumulative analyses supported this pattern across cultural and stimulus contexts. These findings suggest that, in severe cruelty, humanity may be part of what makes suffering worth inflicting.

Keywords: Dehumanization · Humanization · Cruelty · Psychopathy · Sadism · Intimate partner violence · Implicit cognition · Moral psychology

2026 · Manuscrit soumis

The Quality of Replication: What Many Labs 4 Can Teach Us About Large-Scale Science

Chatard, A. (2026). The Quality of Replication: What Many Labs 4 Can Teach Us About Large-Scale Science [Manuscript submitted for publication]. Université de Poitiers.

Lire l’abstract / Résumé

The Quality of Replication: What Many Labs 4 Can Teach Us About Large-Scale Science

Many Labs 4 (ML4; Klein et al., 2022), a large-scale preregistered replication project, reported a failure to replicate a classic finding in social psychology: the mortality-salience effect. In an autopsy of this replication failure, we examined whether the apparent absence of the effect might reflect poorer data quality in smaller laboratory samples. We found that the four laboratories that failed to reach the preregistered minimum recruitment target of 80 participants had more than three times the rate of missing outcome data compared with the other laboratories. Higher rates of missing data were also associated with weaker mortality-salience effects across laboratories. Excluding the four smaller samples yielded positive and statistically significant effects across all three preregistered participant selection procedures used in ML4. A Bayesian multiverse analysis restricted to the larger laboratories also favored a positive effect across all three selection procedures and five prior specifications. These findings suggest that differences in data quality across laboratories may have contributed to the replication failure in ML4. By giving less weight to studies with smaller samples, meta-analyses account for their lower statistical precision, but cannot correct for differences in data quality. Large-scale replication requires not only more data, but better data.

Keywords: Replication · Data quality · Many Labs 4 · Mortality salience · Meta-analysis · Social psychology

2026 · Manuscrit soumis

Chronic pesticide contamination in distributed drinking water and suicide mortality: a nationwide ecological study in France

Chatard, A., Jaafari, N., & Daviller, B. (2026). Chronic pesticide contamination in distributed drinking water and suicide mortality: a nationwide ecological study in France [Manuscript submitted for publication]. Université de Poitiers.

Lire l’abstract / Résumé

Chronic pesticide contamination in distributed drinking water and suicide mortality: a nationwide ecological study in France

Background. Pesticide exposure has been associated with suicidal behaviors, but whether persistent pesticide contamination of distributed drinking water is associated with suicide mortality in the general population remains unknown.

Methods. We conducted a nationwide ecological study across 101 French departments during 2016-2024, comprising 909 department-year observations and combining regulatory drinking-water pesticide measurements with official suicide mortality data. Chronic contamination was defined as the mean annual percentile rank of departmental total-pesticide concentrations. Spatial error models with calendar-year fixed effects related chronic contamination to age- and sex-standardized suicide mortality. Additional analyses examined territorial covariates, alternative indicators of agricultural pesticide pressure, temporal consistency, between- and within-department associations, and effect modification by sex and age.

Results. After accounting for spatial dependence, a 10-point higher chronic pesticide contamination rank was associated with 0.46 additional suicide deaths per 100,000 inhabitants (95% CI [0.13, 0.78], p = .006). Simultaneous adjustment for socioeconomic disadvantage, healthcare availability, and agricultural operators reduced the estimated association by 31%. The association remained significant after separate adjustment for three alternative indicators of agricultural pesticide pressure: pesticide purchasing intensity, cultivated agricultural land share, and crop pressure. Associations were positive in all nine years and primarily reflected persistent between-department differences rather than annual within-department fluctuations. There was no evidence of effect modification by sex, whereas the association increased across ordered age groups (p for interaction < .001), with significant associations from ages 35–44 onward and the largest estimates at ages 75–94.

Conclusions. Chronic pesticide contamination of distributed drinking water was associated with suicide mortality, and this association was not readily explained by territorial characteristics. Moreover, it persisted after adjustment for alternative indicators of agricultural pesticide pressure, indicating some specificity to drinking-water contamination. Finally, the age pattern is compatible with cumulative exposure and/or greater age-related susceptibility, although the ecological design precludes causal inference.

Keywords: Pesticides · Drinking water · Chronic contamination · Suicide mortality · Environmental epidemiology · Public health

2026 · Manuscrit soumis

Implicit suicidal cognition and self-injury: Prediction of suicidal and nonsuicidal thoughts and behaviors over six months

Chatard, A., Landrault, H., Harika-Germaneau, G., Daviller, B., Delbreil, A., Besnier, M., Leblanc, P.-M., & Jaafari, N. (2026). Implicit suicidal cognition and self-injury: Prediction of suicidal and nonsuicidal thoughts and behaviors over six months [Manuscript submitted for publication]. Université de Poitiers.

Lire l’abstract / Résumé

Implicit suicidal cognition and self-injury: Prediction of suicidal and nonsuicidal thoughts and behaviors over six months

Background: Prior research suggests that implicit suicidal cognition, as measured by the Suicide Implicit Association Test, predicts future suicide attempts. Yet suicide attempts represent only one endpoint within a broader spectrum of suicidal and nonsuicidal self-injurious thoughts and behaviors. Many psychiatric patients struggle with these thoughts and behaviors even without making a suicide attempt, making them important targets for risk assessment and prevention.

Methods: This prospective cohort study tested whether implicit suicidal cognition predicts suicidal and nonsuicidal self-injurious thoughts and behaviors over six months. Participants were 203 psychiatric inpatients recruited from two psychiatric hospitals. At baseline, patients completed the Suicide Implicit Association Test, and suicide risk and psychiatric disorders were assessed with the Mini-International Neuropsychiatric Interview. At six months, suicidal and nonsuicidal self-injurious thoughts and behaviors were assessed by telephone or face-to-face clinical interview. Self-injurious thoughts included suicidal ideation, suicide plans, and nonsuicidal self-injury thoughts; self-injurious behaviors included suicide attempts and nonsuicidal self-injury behaviors. All analyses controlled for depressive disorder.

Results: Implicit suicidal cognition was associated with baseline suicide risk, OR = 2.05, 95% CI [1.02, 4.12], p = .044. At follow-up, implicit suicidal cognition was associated with increased risk of self-injurious thoughts, OR = 3.65, 95% CI [1.89, 7.07], p < .001, and self-injurious behaviors, OR = 2.25, 95% CI [1.01, 5.00], p = .046.

Conclusions: These findings extend previous evidence that the Suicide Implicit Association Test predicts future suicide attempts and suggest that implicit suicidal cognition may index a wider vulnerability to suicidal and nonsuicidal self-injurious processes that is not reducible to depressive disorder.

Keywords: Implicit suicidal cognition · Suicide IAT · Suicide risk · Nonsuicidal self-injury · Prospective prediction · Mental health