L'etat de la science en général elle-même

[Rincevent]

(Hat tip à Michael Gundill)

 

Vous connaissez le DSM, la "bible" de la psychiatrie ? Il y a régulièrement de nouvelles éditions, et il semble que dans la nouvelle édition, il y ait quelques conflits d'intérêt financiers. En se limitant aux auteurs américains, la majorité d'entre eux ont reçu pendant la période de rédaction du pognon de la part de l'industrie pharmaceutique, pour un montant total de plus de 14 millions de dollars (il se peut que ça soit davantage, mais c'est ce que les auteurs de cette étude ont réussi à retracer).

 

https://www.bmj.com/content/384/bmj-2023-076902.full

[Lancelot]

Clark, C. J., Fjeldmark, M., Lu, L., Baumeister, R. F., Ceci, S., Frey, K., ...Tetlock, P. E. (2024). Taboos and Self-Censorship Among U.S. Psychology Professors. Perspectives on Psychological Science, 20(5), 941–957.

https://journals.sagepub.com/doi/10.1177/17456916241252085

 

En gros ils ont une série de 10 affirmations qu'ils ont demandé à des chercheurs en psychologie d'évaluer selon différents critères.

 

Quote

1. “The tendency to engage in sexually coercive behavior likely evolved because it conferred some evolutionary advantages on men who engaged in such behavior.”

2. “Gender biases are not the most important drivers of the under-representation of women in STEM fields.”

3. “Academia discriminates against Black people (e.g., in hiring, promotion, grants, invitations to participate in colloquia/symposia).”

4. “Biological sex is binary for the vast majority of people.”

5. “The social sciences (in the United States) discriminate against conservatives (e.g., in hiring, promotion, grants, invitations to participate in colloquia/symposia).”

6. “Racial biases are not the most important drivers of higher crime rates among Black Americans relative to White Americans.”

7. “Men and women have different psychological characteristics because of evolution.”

8. “Genetic differences explain non-trivial (10% or more) variance in race differences in intelligence test scores.”

9. “Transgender identity is sometimes the product of social influence.”

10. “Demographic diversity (race, gender) in the workplace often leads to worse performance.”

 

Vous noterez que toutes sauf la 3 sont contre le sens du vent (qui est très largement gauchiste dans le monde académique).

 

 

Ici vous voyez à quel point les chercheurs déclarent qu'ils ne seraient pas confiant pour partager leur opinion sur ces questions (gauche), à quel point ils pensent que ces opinions sont vraies (milieu) et à quel point ils pensent que la recherche sur ces questions devrait être découragée (droite).

 

Quote

Scholars generally did not want to discourage research. The most discouragement was observed for a genetic contribution to IQ differences, but the mean was still well below the midpoint. This conclusion also contained the most variance, indicating relatively high disagreement about whether this research should be discouraged.

 

C'est déjà assez édifiant, mais si on comparait par sexe ?

 

 

Tiens tiens il y a une grosse fracture qui va exactement dans le sens que vous imaginiez.

 

Quote

Female scholars were more left-leaning (M = 20.86, SD = 16.03) than male scholars (M = 27.90, SD = 18.70), t(401) = 3.93, p < .001, and younger, t(400) = 4.73, p < .001. Conservatism was associated with stronger beliefs that all taboo conclusions are true, rs =.19 to .40, ps < .001. One exception was a strong association between conservatism and disbelief that academia discriminates against Black people, r = −.40, p < .001.

 

 

Quote

With at least small effects, in almost every case, males self-censored more than females (p < .005 for four conclusions), and females wanted to discourage research more than males (p < .005 for seven conclusions; see Table 4 and Figs. S2 and S3 in the Supplemental Material). There were two exceptions: (a) With a small but not significant effect, female professors self-censored more regarding discrimination against conservatives in the social sciences, and (b) male and female professors similarly had almost no desire to discourage research into discrimination against Black people in academia.

 

Je vous passe la figure qui montre que plus les chercheurs adhèrent à des affirmations "de droite", plus ils sont reluctant à les partager, mais regardez ce qu'il en est pour décourager la recherche :

 

Quote

For nearly all taboo conclusions, those who believed the conclusions were false had stronger desires to discourage research, rs = −.09 to −.25, ps < .048 (p < .005 for six). The one exception was again for racial discrimination in academia, but this one was trending in the same direction, r = −.08, p = .097. These associations indicate that professors with more socially desirable beliefs (i.e., beliefs that taboo conclusions are false) have stronger desires to deter research into taboo topics.

 

Alors vous me direz, de quoi ont-ils peur exactement ?

 

Quote

When reporting the consequences they would face if they shared their own empirical beliefs openly, professors were quite concerned about getting attacked on social media (M = 64.48, SD = 33.39), being ostracized by peers (M = 54.80, SD = 33.22), and being stigmatized or labeled pejorative terms (M = 53.34, SD = 34.70). Scholars had low to moderate concerns about disciplinary actions (M = 32.04, SD = 31.96), student boycotts (M = 40.69, SD = 33.62), guilt-by-association harm to students and colleagues (M = 40.89, SD = 33.22), and career-damaging biases against them (M = 44.79, SD = 34.29). These findings suggest that most psychology professors hold some empirical beliefs they consider socially costly.

Scholars were relatively unconcerned about threats of physical violence (M = 26.95, SD = 28.38) and getting fired (M = 17.57, SD = 24.69). [...]  Untenured faculty were not significantly more concerned about getting fired (M = 20.37, SD = 23.91) than tenured faculty (M = 16.83, SD = 24.93), t(417) = 1.28, p = .201, d = 0.14. This lack of difference could be due to a floor effect, although only 40.4% had no concern about getting fired—and recall that this concern is about whether they would get fired if they shared their own empirical beliefs openly, not a concern about sharing some hypothetical and extreme belief. Thus, the majority of professors hold empirical beliefs that they perceive to be sufficiently socially unwelcome that it would increase risk of termination if others were to discover those beliefs.

 

Et n'allez pas croire qu'avoir une position permanente les protège.

 

Quote

There were no differences between the tenured and the untenured on fear of any consequences. We also computed a self-censorship index across all taboo conclusions, ∝ = .92, and found that tenured (M = 40.01, SD = 26.30) and untenured professors (M = 40.55, SD = 25.56) self-censor to virtually identical degrees, t(410) = 0.16, p = .869, possibly because tenure provides no protection against the consequences scholars fear most—ostracism, social-media attacks, and stigmatization.

 

Bon, ça c'était le risque évalué par les chercheurs pour leurs propres opinions. Que pensent-ils des conséquences pour les opinions des autres ?

 

Quote

Women (semipartial r = .19, p < .001) and more left-leaning professors (semipartial r = −.13, p = .006) considered moral concerns a more legitimate reason to retract. And for risks of extremists misconstruing and weaponizing results, only gender predicted retraction support, with women endorsing this reason more (semipartial r = .11, p = .023).

[...]

And women professors (semipartial r = .10, p = .038) and younger professors (semipartial r = −.14, p = .006) more highly endorsed firing professors on the basis of moral concerns about their research conclusions.

 

Mais et la liberté de la recherche là dedans ?

 

Quote

A slim majority of professors (52.3%) reported that scholars should be completely free to pursue research questions without fear of institutional punishment for their conclusions. By contrast,1.6% said scholars should not have this freedom, and 46.0% said it’s complicated. [...] When gender, ideology, and age were simultaneously regressed on support for academic freedom, with small but not always significant effects, female professors (semipartial r = −.10, p = .048), more left-leaning professors (semipartial r = .11, p = .024), and younger professors (semipartial r = .22, p < .001) were less supportive of complete academic freedom.

 

Mais et la vérité ?

 

Quote

A slim majority of professors (56.5%) reported that scientists should prioritize truth when truth and social-equity goals come into conflict. By contrast, 3.1% prioritized social equity over truth, and 40.5% said it’s complicated. [...] When gender, ideology, and age were simultaneously regressed on prioritization of truth, female professors (semipartial r = −.15, p = .002), more left-leaning professors (semipartial r = .23, p < .001), and younger professors (semipartial r = .18, p < .001) were relatively less likely to prioritize truth over social equity.

 

Je ne sais pas vous mais en ce qui me concerne je trouve cette "slim majority" assez glaçante.

 

Donc concrètement quelles actions préconisent ces chercheurs contre des dissidents ?

 

Quote

The last question before our open-ended response asked scholars how much they would support various actions against scholars who draw the most taboo conclusions—those that involve genetic or evolutionary explanations for group differences in socially important outcomes (i.e., in domains in which women or Black people underperform relative to men or White people). Scholars very strongly supported normal scientific criticism, such as commentaries (M = 92.69, SD = 15.96). There was little support for removing such faculty from leadership positions (M = 25.37, SD = 28.75) and disinviting them from talks (M = 22.56, SD = 26.91). There was very little support for socially ostracizing them (M = 14.62, SD = 19.54), publicly labeling them pejorative terms (M = 10.18, SD = 16.35), refusing to publish their work (M = 11.87, SD = 19.04), not hiring or promoting them (M = 14.65, SD = 20.68), stigmatizing their graduate students and coauthors (M = 5.34, SD = 12.28), firing them (M = 6.46, SD = 13.68), or shaming them on social media (M = 11.01, SD = 19.43). Support for normal scientific criticism was either unrelated or negatively related to support for all other punishments (rs = −.10–.02), whereas support for all other punishments were positively related (rs = .36–.72).

 

Ouf, donc la (grosse) majorité supporte ici une réaction saine (critique scientifique uniquement), mais tout de même une minorité aussi grande que 25% pour des sanctions qui dépassent le travail de recherche.

 

Quote

there were no gender differences in support for normal scientific criticism or stigmatizing graduate students and coauthors. However, with at least small effects, women were more supportive than men of ostracism, public labeling with pejorative terms, talk disinvitations, refusing to publish work regardless of its merits, not hiring or promoting even if typical standards are met, terminations, social-media shaming, and removal from leadership positions.

[...]

When gender, ideology, and age were simultaneously regressed on support for all actions with at least small effects, females were more supportive of four punishments, more left-leaning professors were more supportive of eight punishments, and younger scholars were more supportive of eight punishments.

 

Je me dis que j'aurais aussi pu poster ça dans le thread sur le féminisme en fait...

 

Deux points de discussion intéressants :

Quote

Most, but not all, scholars believed that some empirically supported conclusions cannot be mentioned without punishment. Scholars with more socially desirable beliefs were less likely to self-censor, less fearful of punishments, and less likely to perceive the existence of taboos. These patterns may explain why scholars sometimes quarrel about whether academic freedom is at risk—scholars are likelier to notice boundaries when they have crossed them (in their own minds, if not publicly). These patterns of self-censorship also suggest that professional discourse surrounding taboo topics (e.g., at conferences, in faculty meetings, on social media) may be systematically biased toward rejecting taboo conclusions because those who hold taboo empirical beliefs are more likely to remain silent than others.

 

Quote

Most professors were concerned about social ostracization, name-calling, and social-media attacks. This fear might stem from false beliefs about how many scholars endorse such punishments, or it might be a reaction to the small minority that does endorse them. A vocal minority and silent majority may have created a seemingly hostile climate against taboo conclusions and the scholars who forward them, even if the silent majority has great contempt for the vocal minority. Future research should test these possibilities more directly.

 

Pour le second point je serais moins optimiste que les auteurs, leur "small minority" est potentiellement autour de 20% et ne fait que grandir avec la féminisation et les nouvelles générations de chercheurs. Je pense que les chercheurs comprennent extrêmement bien leur environnement et ses risques.

[Marlenus]

4 hours ago, Lancelot said:

 

 

 

Je ne sais pas vous mais en ce qui me concerne je trouve cette "slim majority" assez glaçante.

 

Je la trouve étonnante, dans le sens où pour moi c'est bien haut, mais je parie sur un biais de "Je dis ce qui me semble juste mais j'éviterais de toucher à ce qui est explosif et je ne chercherais la vérité que là où elle ira dans le sens du vent".

Etre le lanceur d'alerte dans un groupe sur le fait que le groupe se trompe est rarement une bonne idée, en tout cas si on veut rester dans le groupe. Ou alors il faut très très bien le préparer.

[Rincevent]

Il y a 2 heures, Marlenus a dit :

Etre le lanceur d'alerte dans un groupe sur le fait que le groupe se trompe est rarement une bonne idée

... d'un point de vue égocentré. Quand tu veux le bien d'une personne, tu lui dis la vérité ; quand tu ne cherches que ton bien à toi, tu lui dis ce qu'il souhaite entendre.

[Marlenus]

39 minutes ago, Rincevent said:

... d'un point de vue égocentré. Quand tu veux le bien d'une personne, tu lui dis la vérité ; quand tu ne cherches que ton bien à toi, tu lui dis ce qu'il souhaite entendre.

Il faut avoir un certain poids dans le groupe pour que cela ait potentiellement un effet. Si tu es un noname du groupe, tu te retrouveras juste avec ta morale. C'est important pour certains, mais certainement pas pour plus de 50% de la population au prix de sacrifier sa carrière et ses relations sociales.

 

Perso j'assume, pour l'avoir vécu, être parti plutôt que de m'être battu. Ce que j'ai perdu en intégrité morale, je l'ai gagné autre part.

D'ailleurs partir du groupe pour éviter la confrontation avec le groupe est un choix qui est beaucoup plus courant que de s'opposer frontalement quand tu sens que tu ne vas pas être soutenu.

 

[Rincevent]

il y a 9 minutes, Marlenus a dit :

D'ailleurs partir du groupe pour éviter la confrontation avec le groupe est un choix qui est beaucoup plus courant que de s'opposer frontalement quand tu sens que tu ne vas pas être soutenu.

Oh je le sais bien, crois-le.

 

Mais il y a une différence entre partir du groupe, et y rester en fermant ta gueule voire en disant le contraire de ce que tu penses vrai (et c'est cette dernière situation que le papier cité étudie).

[Marlenus]

Just now, Rincevent said:

Mais il y a une différence entre partir du groupe, et y rester en fermant ta gueule voire en disant le contraire de ce que tu penses vrai (et c'est cette dernière situation que le papier cité étudie).

Ceux qui sont parti, n'ont pas répondu au questionnaire (si j'ai bien suivit).,donc il y a un biais de sélection.

Faudrait voir combien quittent la recherche à cause de cette ambiance et vu ce qui est décrit, je doute que ce soit négligeable.