Last week’s post highlighted a paper by Casadevall and Feng which argued that the culture of science is in urgent need of reform. In particular, they argued, “incentives in the current system place scientists under tremendous stress, discourage cooperation, encourage poor scientific practices”. I agreed with the thrust of the article, and I wondered how widespread these views might be, so I set up a survey to poll readers on each point raised in the original article.
At the time of writing there were 92 respondents – clearly a smallish, self-selecting sample, but a reasonable number from which to assess the degree of consensus and compare support for the different points. There was overall agreement with the diagnosis offered by Casadevall and Feng – between 76% and 97% of respondents agreed with each point. Views on the prescription were more nuanced, but again there was overall support for each recommendation (ranging between 61% and 92% of respondents). The survey is still open – I’ll report back if/when I have 1000 responses, so do feel free to share the link.
Casadevall and Feng argued that the scientific careers structure was imbalanced, resembling a pyramid scheme. 97% of respondents agreed that and 90% felt that this was problematic.
They claim that that targets and competition are distorting science. Again 88% agreed and 80% felt that this was problematic.
Similarly, 90% agreed that the importance of priority encourages citation bias, secrecy and the appropriation of others’ ideas and data and 86% felt that this was problematic.
There was also fairly strong support for the claim that credit is disproportionatly awarded to PIs for what is truly the product of teamwork. Here 77% of respondents agreed, but 7% felt that this was not problematic and 5% disagreed.
They argued that winner-takes-all competition encourages cheating and fraud; 89% of respondents agreed and 82% felt that this was problematic.
There was almost universal agreement, 97%, that pressure to publish (rather than discover, understand and explain) is leading to inefficiencies.
76% agreed or strongly agreed with the recommendation that promotion criteria be revised to incorporate peer evaluations reflecting scientific quality of research and the specific contribution of authors.
88% agreed with the recommendation that training in probability and statistics be improved and that application of statistical principles should be more consistent.
Views were more evenly divided on recommendations concerning the importance of philosophy in science. 64% agreed that training should be improved and that philosophy should take a more central role, but a substantial minority 16% actively disagreed with the proposal and 18% were neutral.
There was some support (61%) for the idea that checklists should be developed to avoid basic errors, but 30% of respondents were neutral and 9% disagreed with the proposal.
68% agreed with that the scientific culture is too “masculine”, placing excessive emphasis on competition, encouraging conflict and discouraging colaboration, 24% of respondents were neutral and 7% disagreed.
There was very strong support for the recommendation that generosity and cooperation should be recognized and rewarded (92% agreed, 64% strongly). Only 8% were neutral on this issue and not one respondent disagreed.
There was also strong support (86% agreed, 57% strongly) for the recommendation that reforms should focus on developing supportive and productive groups.
From this preliminary survey it seems that may be little dissent, from Casadevall and Feng’s critique of the scientific culture, and support for many of the reforms they advocate. It seems to me that if they are right these are very important issues that scientists must address. We must ask ourselves, what is stopping it from happening? The article begins with a very aposite quote from Machiavelli: “… He who innovates will have for enemies all those who are well off under the existing order of things, and only lukewarm supporters in those who may be better off under the new.”
The survey is still open.
Google Forms (part of Google Docs) are really easy to use; highly recommended.