A 2012 paper<\/a> showed there was indeed a simple intervention: in both lab tests and field studies, the researchers found that having people sign an honesty pledge<\/strong> at the top of the form caused a significant increase in honest answers. The most prominent author listed was Dan Ariely<\/strong>, author of the global bestseller Predictably Irrational<\/em><\/a>. Ariely wrote an entire book about the topic of dishonest behavior, The (Honest) Truth About Dishonesty<\/em><\/a>, also published in 2012.<\/p>\n In 2020, a paper<\/a> by seven authors that included the original five found that the original findings couldn\u2019t be replicated. The new paper said, \u201cThe current paper updates the scientific record by showing that signing at the beginning is unlikely to be a simple solution for increasing honest reporting.\u201d This paper noted that some government agencies had adopted the \u201csign first\u201d approach based on the original research.<\/p>\n \u201cUpdating the science record\u201d didn\u2019t go far enough. In 2021, the original paper was retracted<\/a>. after detailed analysis of its data by the blog Data Colada found \u201cevidence of fraud.\u201d<\/a><\/p>\n A report today from NPR<\/a> has resurfaced this problematic research by publishing a letter from The Hartford<\/strong>, an insurance company. The firm\u00a0provided data on about six thousand vehicles, but the published study described a data set of more than three times<\/strong> that number. According to an analysis by The Hartford, \u201c\u2026it is clear the data was manipulated inappropriately and supplemented by synthesized or fabricated data.\u201d The letter goes on to detail why they drew that conclusion using both statistical and, oddly, typographic analysis. The apparently bogus data is in a different font than the original provided data.<\/p>\n Taking Ariely at his word, this messy situation raises again the question of responsibility of listed authors on scientific papers. The paper in question has a modest number of authors \u2013 a mere five, actually below average. The average number of authors on scientific papers grew from two in 1980 to seven in 2019.<\/p>\n With most papers having multiple authors, what is the responsibility of each author to verify all data, methodology, etc.? Is every author responsible? Is the first author<\/a> the one who assumes full responsibility? What does being listed as an author indicated about the individual contribution?<\/p>\n Being listed as an author on a paper for an incidental contribution is usually a good thing \u2013 publications are the lifeblood of academic success. When things go wrong, of course, being an author becomes a liability.<\/p>\n At this point, Ariely likely wishes he had been listed in the original paper\u2019s acknowledgements rather than as an author. He wasn\u2019t the primary author, but, as the most famous name in the author list, he becomes the headline.<\/p>\n Nobody knows better than a behavioral scientist that denying a false claim can reinforce the belief<\/a> being refuted. The more traction the story gets, the more Ariely\u2019s reputation can become tarnished \u2013 even if he had nothing to do with the apparently fabricated data.<\/p>\n This questionable paper is part of a much larger issue in the social sciences: studies often find significant effects for interventions, but other researchers are unable to replicate them. This has been dubbed \u201cThe Replication Crisis\u201d by some.<\/p>\n I spoke with Ariely<\/a> about this topic in 2017. He downplayed the \u201ccrisis\u201d concept, noting that many attempts to replicate studies vary in some important way from the original research. The subjects differ in things like age, geography, culture, and other demographic factors. Research methodology and sample size can differ. Differing results are to be expected when the replication isn\u2019t identical.<\/p>\n Ariely said the welcomed replication studies, particularly those that attempt to expand the learning from the original research, such as identifying conditions would make the result stronger or weaker.<\/p>\n Ariely also urged caution on accepting the results of one study:<\/p>\n You see one experiment, don\u2019t get convinced. No matter what the experiment is, don\u2019t be convinced 100%. Change your belief a little bit in the direction of the data.<\/p>\n<\/blockquote>\n Some replication problems, of course, come from bad research. Under pressure to publish important findings, researchers can torture the data they collect until something significant emerges. Outlying data points can be discarded as errors to produce a stronger result. Conclusions based on a small number of subjects can be expressed as a general understanding of human behavior.<\/p>\n Less common is wholesale fabrication of large data sets as is claimed to have happened here.<\/p>\n In the case of the \u201csignature at top\u201d paper, the problem may lie, at least in part, with co-author Francesca Gino<\/strong>, a behavioral scientist at Harvard Business School. Last month,the Data Colada bloggers<\/a> say they found evidence of fraud in four<\/strong> of her papers. Her status at Harvard is currently \u201con administrative leave.\u201d<\/p>\n![\"woman](\"https:\/\/www.neurosciencemarketing.com\/wp-content\/uploads\/2023\/07\/woman-nose.jpg\")
Business leaders love behavioral science interventions that are simple, free, and change behavior in a significant way. Nudging people to save more for retirement by making plans opt-out rather than opt-in has helped millions of employees have a more secure future. Businesses collect all kinds of information using forms, like employment applications, market surveys, and so on\u2026 Could a simple intervention<\/strong> make people more honest when completing a form?<\/p>\nThe Problem with the \u201cHonesty\u201d Paper<\/h2>\n
Dan Ariely Disclaims Responsibility For Fabricating Data<\/h2>\n
![\"Growth](\"https:\/\/www.neurosciencemarketing.com\/wp-content\/uploads\/2023\/07\/paper-authors-240x181.png\")
<\/a>Ariely denies being involved in any data fabrication. He told NPR, \u201cGetting the data file was the extent of my involvement with the data.\u201d<\/p>\nThe Backfire Effect<\/h2>\n
Replication in Behavioral Science<\/h2>\n
\n
Bad News for Behavioral Science<\/h2>\n