Admettre ouvertement une erreur grave vs. la dissimuler ou l'esquiver
Si vous agissez
Reconnaître ouvertement la faute grave
14%
Si vous n’agissez pas
Dissimuler ou détourner l'attention de la faute grave
40%
Pourcentage de personnes qui regrettent ensuite chaque choix. Les barres et le registre complet s’affichent ci-dessous.
Mode de vie
Dernière révision 2026-05-13
Qualité des preuves 3.38/5
Score d’évaluation en huit dimensions selon la
grille de qualité
. Chaque dimension notée de 1 à 5.
D1 Vérification des sources
3/5
D2 Autorité et indépendance des sources
3/5
D3 Précision du taux de regret
2/5
D4 Comparabilité des sources
2/5
D5 Motif de Gilovich
4/5
D6 Qualité de la prose
4/5
D7 Complétude des réserves
5/5
D8 Qualité de l’échantillon
4/5
Moyenne3.38/5
Données de substitution — aucune enquête directe sur les regrets n'existe pour cette décision. Les taux sont dérivés des scores de satisfaction et des obstacles d'accès plutôt que de questions portant directement sur les regrets. Voir les mises en garde ci-dessous.
Regret d'action
Reconnaître ouvertement la faute grave
14%
aucune enquête directe ne le mesure ; les ~14% sont un indicateur indirect estimé par l'auteur pour la minorité dont l'aveu se retourne contre elle (préjudice de carrière ou de relation), ancré sur les données de sens de l'effet montrant qu'admettre sa faute est largement valorisé et propice à la restauration de la confiance — et non un taux mesuré de personnes ayant avoué et rapportant de mauvais résultats
Professionnels et adultes ayant révélé une erreur significative
enquête transversale, 2015 ; expérimentale, 2004
Regret d'inaction
Dissimuler ou détourner l'attention de la faute grave
40%
aucune enquête directe ne le mesure ; les ~40% sont un indicateur indirect estimé par l'auteur, ancré sur le constat du World Regret Survey selon lequel les regrets moraux (qui incluent la dissimulation d'erreurs) figurent parmi les types de regret les plus intenses et persistants — et non un taux mesuré de personnes ayant dissimulé et rapportant une culpabilité durable
Adultes ayant sciemment caché une erreur personnelle ou professionnelle significative
estimations cliniques, 2021 ; enquête à grande échelle sur le regret, 2021
% regrettent ce choix
Reconnaître ouvertement la faute graveDissimuler ou détourner l'attention de la faute grave
14%40%
inaction dominates — L'inaction domine — la plupart regrettent de ne pas avoir agi.
Décisions associées
Décisions sémantiquement similaires — même terrain, compromis différents.
The peer-reviewed and survey evidence consistently points to admitting a serious mistake producing positive outcomes for most people in most contexts. Ma and colleagues’ experimental trust-game study (Frontiers in Psychology, 2019) found that receiving an apology — an admission of fault — made participants more likely to trust their counterpart than receiving no apology, an effect mediated by perceived trustworthiness; the benefit held for those experiencing less negative emotion and weakened for those still highly aggrieved. A Dale Carnegie Training survey of nearly 3,100 employees across 13 countries (reported in The Public Manager) found that 81 percent of respondents valued a leader who admits when they are wrong, while only 41 percent felt their supervisors could be trusted to do so consistently — the largest importance-versus-practice gap of any leadership behavior surveyed. Neither source measures how often admission backfires on the person who admits, so the ~14 percent action rate shown here is not a measured consequence rate but an author proxy: a small-minority estimate for the subset of disclosures made in hierarchical, punitive, or high-emotion environments where admission is weaponised against the person who made it. It is flagged proxy_only and should be read as a directional estimate, not a survey statistic.
The inaction side has no direct survey: no published study reports the proportion of people who concealed a serious mistake and later report ongoing guilt or regret about the concealment itself. The ~40% rate shown here is an explicit author proxy (the entry is flagged proxy_only), anchored only on direction-of-effect evidence. Daniel Pink’s World Regret Survey (n=26,000+, 105 countries) is the strongest anchor: moral regrets — the category that includes covering up mistakes and acting against one’s own values — make up roughly 10 percent of all reported regrets but are rated among the most emotionally intense and long-lasting of all regret types. Secrecy research synthesised by Slepian (Current Directions in Psychological Science, 2024; ~1,000 participants, thousands of secrets) finds that the more a person’s mind wanders to a concealed matter, the more it harms well-being, with concealment associated with lower life satisfaction, anxiety, and isolation. Both lines of evidence point to concealment carrying a sustained psychological cost, but neither establishes a rate; the 40% figure should be read as a directional estimate, not a measured survey statistic.
The regret_delta of -0.26 reflects the pattern identified by Gilovich and Medvec across multiple decision domains: in the short term, admission is the more painful option — the immediate exposure, embarrassment, and risk of negative consequences are concrete and salient, while the long-term cost of concealment is diffuse and delayed. Over time, this asymmetry reverses. People who admitted mistakes and survived the consequences generally move past them; people who concealed significant errors carry the ongoing psychological burden of concealment and the escalating risk of discovery. The Gilovich inaction-dominance pattern fits this entry well. The main caveat is context: in highly hierarchical or legally consequential environments, where admission carries career-ending or criminal risk, the action-regret rate can approach or exceed the inaction-regret rate, and the entry’s inaction-dominance finding should not be applied to those high-stakes outlier situations without accounting for the specific professional and legal context.
Sources : action
Registre des sources
Chaque chiffre ci-dessous correspond à ce que la source a rapporté, avec la citation textuelle sur laquelle nous nous sommes appuyés et la méthode de calcul. Cliquez sur un lien pour vérifier directement.
2/2 sources vérifiées de manière indépendante, mot pour mot, par rapport à la source citée
[1]The Public Manager (Association for Talent Development), reporting Dale Carnegie Training survey — The Value of Acknowledging Your Mistakes
Vérifié
Source de référence
In a Dale Carnegie Training survey of nearly 3,100 employees across 13 countries, 81% said having a leader who admits when they are wrong is important or very important to inspiring their best efforts, but only 41% said their supervisors could be trusted to do so consistently -- a 40-point gap, the largest of any leadership behavior surveyed.
Extrait
“A recent online survey conducted by Dale Carnegie Training of nearly 3,100 employees across 13 countries found that "admitting when they are wrong" produced the largest gap of any of the leadership behaviors in terms of the difference between its importance to employees and its consistent performance by supervisors. Eighty-one percent of respondents said that having a leader who will admit to being wrong is important or very important to inspiring them to give their best efforts at work, but only 41 percent said their supervisors could be trusted to do so consistently -- a gap of 40 percent.
”
Données source de
2016-01-01
Consulté le
2026-07-01
Vérification
Extrait récupéré et confirmé de manière indépendante, mot pour mot, par rapport à la source citée lors de notre audit de vérification.
Calcul
Dale Carnegie Training survey, n=~3,100 across 13 countries, reported in The Public Manager (Association for Talent Development). This source measures the VALUE employees place on leaders admitting mistakes: 81% say it is important, but only 41% see it consistently -- a 40-point importance-vs-practice gap. CRITICAL: it does NOT measure the proportion of people who admitted a serious mistake and later faced negative career consequences or regretted the admission. It cannot anchor a specific action-side regret rate. It supports only the DIRECTION -- admission is broadly valued, and a majority of employees would rather work under leaders who admit error -- consistent with a low action-side regret rate, but the ~14% magnitude is an author proxy disclosed in the regret_display, not a figure from this survey. Replaces a prior citation to a Workstars blog URL that carried a "14% faced lasting negative career consequences" figure the source does not contain (that consequence rate could not be located in any published survey).
[2]Frontiers in Psychology (PubMed Central), Ma et al. 2019 — Apologies Repair Trust via Perceived Trustworthiness and Negative Emotions
Vérifié
Revue par les pairs
Experimental trust-game study (PMC6457316): participants who received an apology (an admission of fault) were more likely to trust their counterpart than those who did not, mediated by perceived trustworthiness; the benefit held for those with less negative emotion.
Extrait
“Participants who received an apology were more likely to trust in their counterpart, compared to those who did not receive an apology, and this relationship was mediated by perceptions of the transgressor's trustworthiness. Additionally, the relationship between apologies and perceptions of the transgressor's trustworthiness was moderated by negative emotions; apologies only improved perceptions of trustworthiness for participants who experienced less negative emotions.
”
Données source de
2019-04-03
Consulté le
2026-07-01
Vérification
Extrait récupéré et confirmé de manière indépendante, mot pour mot, par rapport à la source citée lors de notre audit de vérification.
Calcul
Ma, Wylie, Luo, He, Jiang, Zhang, Xu and Evans (2019), Frontiers in Psychology, PMC6457316. Experimental trust-game study on trust repair after a violation. Cited as corroborating authoritative source for the action side: admitting fault (apologizing) generally repairs trust and improves perceived trustworthiness relative to not apologizing -- but the effect is dampened when the wronged party holds strong negative emotion. This supports the DIRECTION that admission is broadly beneficial (low action-side regret) while noting a moderating condition (high-emotion / adversarial contexts) where admission helps less. It does NOT provide a percentage of admitters who regret admitting, so it supports the direction only, not the ~14% proxy magnitude. Replaces a prior citation to Lozano and Laurent (2019, PLoS ONE, PMC6405044): that paper is real but its actual finding concerns whether observers expect OTHER unrelated agents to shift blame, and it does not support the excerpt previously attributed to it.
Sources : inaction
Registre des sources
Chaque chiffre ci-dessous correspond à ce que la source a rapporté, avec la citation textuelle sur laquelle nous nous sommes appuyés et la méthode de calcul. Cliquez sur un lien pour vérifier directement.
1/2 sources vérifiées de manière indépendante, mot pour mot, par rapport à la source citée
[1]Current Directions in Psychological Science (Slepian, 2024) — The New Psychology of Secrecy
Vérifié
Revue par les pairs
Across studies of ~1,000 participants holding thousands of secrets, the more frequently people's minds wander to a concealed secret, the more that secret harms their well-being -- mind-wandering to secrets is related to lower life satisfaction, anxiety, and loneliness. (The research reports correlational associations, not a percentage of concealers who regret concealing.)
Extrait
“The more frequently people's minds wander to their secrets (outside of concealment contexts), the more those secrets harm their well-being. Mind wandering to secrets increases feelings of shame, isolation, uncertainty, and inauthenticity and is related to lower life satisfaction, anxiety, and loneliness. Frequent concealment of secrets is not by itself associated with stress; the harm comes from repetitive thinking about the concealed information rather than the act of concealment during conversations.
”
Données source de
2024-06-01
Consulté le
2026-06-30
Vérification
Extrait récupéré et confirmé de manière indépendante, mot pour mot, par rapport à la source citée lors de notre audit de vérification.
Calcul
Slepian (2024), "The New Psychology of Secrecy," Current Directions in Psychological Science, Vol. 33 Issue 3, synthesizing the secrecy research program (Slepian et al. 2017 onward, ~1,000 participants, thousands of secrets). Cited as authoritative evidence that concealing a significant matter is associated with sustained psychological burden (lower life satisfaction, anxiety, isolation) via repetitive mind-wandering. CRITICAL: this body of work reports correlational associations between secrecy and well-being -- it does NOT report a percentage of people who concealed a mistake and report ongoing guilt or regret. It cannot anchor a specific rate. It supports only the DIRECTION (concealment carries ongoing cost), not the 40% proxy magnitude, which remains an author estimate disclosed in the regret_display. Replaces a prior Psychology Today citation that had been attributed a "40-60% longitudinal clinical" figure the source does not contain (the source is qualitative clinical commentary with no percentages).
[2]Daniel H. Pink / worldregretsurvey.com — World Regret Survey↗ 3 other entries
Étude primaire
n=26,000+ respondents across 105 countries: moral regrets -- acting against one's own values, including covering up mistakes -- constitute approximately 10% of all regrets but are rated among the most emotionally intense and long-lasting of all regret categories.
Extrait
“[Paraphrase from World Regret Survey data and Pink, The Power of Regret (2022).] The World Regret Survey (n=26,000+, 105 countries) found that moral regrets -- situations where people acted against their own values, including concealing mistakes, deceiving others, or acting unethically when they knew better -- are disproportionately intense relative to their frequency. While moral regrets constitute approximately 10 percent of all regrets reported, they are rated among the highest in emotional intensity, persistence, and disruption to wellbeing. Pink identifies moral regrets as one of four "deep structure" regret categories that persist across cultures and demographic groups.
”
Données source de
2021-01-01
Consulté le
2026-05-13
Calcul
Daniel H. Pink, World Regret Survey (2021), n=26,000+ across 105 countries. Primary authoritative anchor for the inaction-side proxy: moral regrets (the category that encompasses covering up serious mistakes) constitute roughly 10% of all reported regrets but are rated among the most emotionally intense and long-lasting of all regret types. CRITICAL: the survey does NOT measure "what proportion of people who concealed a mistake regret it" -- that is a category-share and intensity finding, not a rate. Moral regrets as a category are the proxy for direction (concealment regret, once present, tends to be severe and persistent rather than transient); they do not establish the ~40% magnitude, which is an author estimate disclosed as a proxy in the regret_display. The prior derivation that blended a non-existent "40-60% clinical midpoint" with this survey to manufacture 51% has been removed.
Réserves
Le taux de regret-action de ~14% est un indicateur indirect explicitement estimé par l'auteur (proxy_only), ET NON un taux mesuré : aucune enquête publiée ne rapporte la proportion de personnes ayant avoué une erreur grave et ayant ensuite subi des conséquences négatives durables ou regretté leur aveu. La seule enquête directe disponible (Dale Carnegie, n=~3 100, 13 pays) ne mesure que la VALEUR que les employés accordent au fait que les dirigeants admettent leurs erreurs — 81% disent que c'est important, 41% le constatent régulièrement — et non un quelconque taux de conséquences ; et les travaux expérimentaux (Ma et al. 2019) montrent que l'aveu restaure largement la confiance, atténué seulement lorsque la partie lésée éprouve une émotion négative forte. Tous deux ancrent le SENS (l'aveu est largement bénéfique, le taux de retour de bâton est donc faible), mais aucun n'établit l'amplitude de 14%, qui est une estimation d'auteur pour une petite minorité de contextes hiérarchiques, punitifs ou à forte charge émotionnelle et peut sous-estimer le taux réel dans les cultures professionnelles les plus punitives. Une version antérieure de cette entrée attribuait un chiffre de « 14% ayant subi des conséquences de carrière négatives durables » à une page de blog Workstars qui ne le contient pas, et citait un article expérimental sans rapport (Lozano & Laurent 2019, sur l'attente que D'AUTRES agents rejettent le blâme) avec un extrait fabriqué ; les deux ont été remplacés par des sources dont les extraits sont mot pour mot. Le taux de ~40% du côté inaction est un indicateur indirect explicitement estimé par l'auteur (proxy_only), ET NON un taux mesuré : aucune enquête publiée ne rapporte la proportion de personnes ayant dissimulé une erreur grave et rapportant une culpabilité ou un regret durable au sujet de cette dissimulation. L'indicateur est ancré uniquement sur des données de sens de l'effet — le constat du World Regret Survey selon lequel les regrets moraux (qui incluent la dissimulation d'erreurs) figurent parmi les types de regret les plus intenses et persistants, et la recherche sur le secret (Slepian) montrant que la dissimulation est associée à un moindre bien-être, à l'anxiété et à l'isolement. Tous deux ancrent le SENS (la dissimulation comporte un coût durable), non l'amplitude. Une version antérieure de cette entrée dérivait un chiffre de « 51% » en mêlant une statistique inexistante de « 40-60% clinique longitudinal » (attribuée à tort à un commentaire de Psychology Today ne contenant aucun tel chiffre) au constat qualitatif d'intensité du World Regret Survey ; ce composite a été supprimé et le côté requalifié en indicateur indirect divulgué. Le contexte importe considérablement : dans les environnements très hiérarchiques, politiquement chargés ou lourds de conséquences juridiques, l'aveu comporte un risque de carrière plus élevé que ne le laissent entendre les données de sens de l'effet largement positives (enquête Dale Carnegie, Ma et al. 2019), et le taux de regret-action dans ces contextes peut approcher ou dépasser le taux de regret-inaction. Le constat de dominance de l'inaction s'applique le plus fortement aux contextes entre pairs et interpersonnels ; dans les environnements d'entreprise à gestion stricte de la performance, le calcul du résultat peut s'inverser. La décision n'est pas non plus toujours nettement binaire : beaucoup de gens admettent des versions partielles de leurs erreurs, se dérobent partiellement, ou révèlent à certains publics mais pas à d'autres, ce qui rend le cadrage action/inaction quelque peu stylisé par rapport à la distribution réelle des comportements.
