Article révisé par les pairs
Résumé : In late 2017, it has surfaced that 10 papers authored by Nicolas Guéguen (NG) suffered from flaws that could be attributed to negligence or possibly scientific misconduct. These concerns appeared after two scholars, Nicholas Brown and James Heathers published extensive analyses of this work on their blogs (cf. O’Grady, 2017 for a summary and Brown & Heathers, 2017 for a full report). This led the chief editors of IRSP, Céline Darnon (CD) and Olivier Klein (OK) to request an investigation from Hans IJzerman (HIJ) about six papers published in the IRSP by Nicolas Guéguen (NG) and IRSP committed to following HIJ’s recommendations: • Dufourcq-Brana, M., Pascual, A., Guéguen, N. (2006). Déclaration de liberté et pied-dans-la-porte. Revue Internationale de Psychologie Sociale, 19, 173–187. • Guéguen, N., & Martin, A. (2008). L’effet de l’imitation sur l’évaluation d’autrui: Une expérimentation dans un contexte de séduction. Revue Internationale de Psychologie Sociale, 21, 5–24. • Guéguen, N. (2011). Mr. Lemaître is a good teacher: An experimental study personological inference activated by names. Revue Internationale de Psychologie Sociale, 24, 107–116. • Guéguen, N., Pascual, A., & Lader, S. (2003). L’effet irradiant d’un engagement sur un comportement non explicitement sollicité: Une évaluation en situation naturelle. Revue Internationale de Psychologie Sociale, 16, 125–130. • Guéguen, N., Jacob, C., & Legohérel, P. (2003). Communication médiatisée par ordinateur et sollicitation à une requête: Une évaluation de l’efficacité de la technique du « Pied-dans-la-Porte» lors d’une interaction par e-mail ou sur un site Web. Revue Internationale de Psychologie Sociale, 16, 125–156. • Pascual, A., & Guéguen, N. (2002). La technique du « vous êtes libre de…»: induction d’un sentiment de liberté et soumission à une requête ou le paradoxe d’une liberté manipulatrice. Revue Internationale de Psychologie Sociale, 15, 51–80. It is worth noting that the present investigation is independent from the investigation conducted by NG’s institution (Univ. de Bretagne Sud). This investigation followed the guidelines of the Commission on Publication Ethics (2009). HIJ has appointed four (independent) research methodologists to examine the manuscripts on the basis of their analyses reports. (i.e., the “statistical investigation”). Separately, HIJ has conducted an investigation into the physical evidence supporting the execution of the studies. This includes requesting the data files, any evidence that the studies were conducted, and the likelihood that the studies could have been conducted as planned (i.e., the “forensic investigation”). As part of the forensic investigation, he contacted NG to help gather the evidence for the studies. NG’s co-authors were contacted to inform them of the procedure, they were informed of problems (if any) in the published research, and they received an opportunity to respond to the report before any action was taken. Alexandre Pascual (AP) answered the questions concerning the paper he co-authored but NG did not. Finally, before the final report was shared with NG, a leading scholar in replicability issues with extensive editorial experience, was asked to comment on the report. The report showed that there were errors in all of the papers. For four of the articles (the first four mentioned above), the extent and nature of the errors suggest that the findings may be unreliable. For two of those (Guéguen, 2008 and Gueguen & Martin, 2011), the report raised suspicions of misconduct, defined here as “the provision of incorrect information in a scientific context of substantial significance, either intentionally or grossly negligently” (Bornmann, 2013, pp. 88–89). Consequently, the report recommends an erratum for two papers (Guéguen, Jacob, & Legohérel, 2003; Pascual & Guéguen, 2002), a statement of concern for two others (Dufourcq-Brana, Pascual, & Guéguen, 2006; Guéguen, Pascual, & Lader, 2003) and the retraction for the two papers mentioned above (Guéguen, 2011; Guéguen, & Martin, 2008). It was our responsibility to make a decision based on the report. Before doing so, we consulted the editorial team of the journal. In addition, we have decided to give NG a final opportunity to respond and comment the report, which he did. Note that we do not include his response to the report as he has not consented to share it publicly. We also asked AP to write an erratum on the 6th paper of the list (Pascual & Guéguen, 2002) and NG to write an erratum on the 5th paper of the list (Guéguen et al., 2003). The erratum AP provided will be published separately (see: Pascual 2019). NG did not provide an erratum for the paper by Guéguen et al. (2003). We believe the investigation has been conducted competently and professionally and are very grateful to HIJ and the experts in data analysis he consulted for their remarkable work. Besides, it is our view (and HIJ’s) that the responses provided by NG do not allow to dispel most of the concerns raised by the report. However, taking into consideration the elements described below, we have decided to deviate from our initial commitment to follow the recommendations, for the following reasons: 1. Although he had not answered to former requests sent by HIJ, NG has responded to our final request to comment the report. 2. The six articles were published a long time ago (between 2002 and 2011). Errors are common especially for papers from this period (Nuijten, Hartgerink, van Assen, Epskamp, & Wicherts, 2016). The report concludes misconduct as per the definition provided above. However, the standards for conducting and evaluating research have evolved since NG published these articles and for that reason, we instead believe it is difficult to establish with sufficient certainty that scientific misconduct has occurred. Hence, we decided not to change the status of one paper and express a concern for five others. None of these papers will be retracted. This status is amenable to change should we receive additional information regarding these papers. The present expression of concern was submitted to NG and coauthors. Some of them (Alexandre Pascual) agreed with its content. Others (Céline Jacob and Patrick Legoherel) stated that they did not. NG and the other co-authors did not respond. It is worth noting that the journal does not now have the same publisher as when these papers were published. Thus, the previous publisher (Presses Universitaires de Grenoble) is in charge of linking these expressions of concern and the erratum to the original publications. In sum, here is our decision regarding these 6 papers (Table Presented).