Scientific Integrity and Self-Interest in Science |
|
| Session Organisers | Miss Julia Jerke (University of Zurich) Mr Justus Rathmann (University of Zurich) |
| Time | Friday 19th July, 13:30 - 14:30 |
| Room | D22 |
Social norms, social control and sanctions are an essential tool to generate social order. According to the Mertonian principle, social norms preserve the advancement of knowledge and the ideal of truth seeking in science. Therefore, researchers are expected to subordinate their personal interest to the ideal of truth seeking science. Yet, these collective goods often conflict with researchers’ individual aspirations for reputation, peer-recognition and career advancement.
In recent years, prominent fraud cases such as Diederik Stapel, Jan Hendrik Schön or Joachim Boldt were uncovered, showing that common rules of social and self-control failed and raising the scientific community’s awareness for fraud and manipulation in science. Often, the self-healing powers of science are stressed. However, it is likely that these cases are not only a few bad apples in an otherwise healthy scientific system. These cases may be symptoms of a ‘publish or perish’ system, where researchers face a high pressure to present original, novel and innovative contributions and, furthermore, a highly incentivized publication market with citations as the coveted currency and exchange rates increasing with journal impact factor. Such a system may generate incentives for questionable research practices (QRP) and scientific misconduct. Also, it misleads honest research such that publishers and reviewers may prefer manuscripts reporting novel and significant results and that researchers, in anticipation, do not even submit manuscripts with negative or insignificant results in the first place. Such selection strategies bias the published scientific literature by overly significant and hypothesis-confirming results, misleading or at least slowing down the advancement of knowledge.
There are already approaches to overcome this dilemma such as preregistration of empirical studies, results-free reviews and changing publishing policies of journals. However, it is still up for evaluation how effective these approaches are or can be in the future.
This workshop welcomes empirical or theoretical contributions addressing QRP in general, scientific misconduct, publication bias, the related file drawer effect, methods to detect and prevent scientific misconduct and related topics.
This includes research that
- discusses and investigates features of the scientific system that may promote misconduct,
- explores the efficiency and efficacy of the review and replication system as a form of organized skepticism,
- assesses the prevalence of QRP and scientific misconduct,
- develops and/or tests methods to detect scientific misconduct, or
- deals with attempts or solutions to overcome the dilemma of a biased scientific literature.
Keywords: scientific integrity, publish or perish, questionable research practices, scientific misconduct, publication bias
Dr Elena Damian (University of Leuven) - Presenting Author
Professor Bart Meuleman (University of Leuven)
Professor Wim van Oorschot (University of Leuven)
In recent years there have been increasing evidence of lack of transparency in scientific woks as well as constant debates about possible ways to achieve greater research transparency (e.g., transparency guidelines, preregistration plans). Academic journals are considered to be one of the actors of the academic community that have a strong and direct influence on the content of an article, and consequently its level of transparency. For example, editorial boards have the capacity to encourage or adopt transparency reporting guidelines, offer the option to provide additional materials in the form of online supplementary materials, or require full replication materials. Given that articles published in high-prestige journals receive more citations and are considered to have a higher methodological quality, with this paper, we are interested in finding out if these journals publish more transparent studies as well. More specifically, we examine the association between several journal as well as article and first author characteristics (e.g., journal impact factor, number of issues per year, article length, author employment status) and article evaluation and replicability transparency index (Damian, Meuleman, van Oorschot, forecoming). Evaluation transparency refers to the question whether readers have sufficient information to assess the validity and reliability of a study’s findings. Replicability transparency regards reporting all the necessary information for other to be able to conduct a direct replication. These associations are tested using detailed data about 305 cross-national studies published, between 1986 and 2016, in one of 29 social sciences journals.
Professor Juan Castillo (Departamento de Sociología - Universidad de Chile) - Presenting Author
The notions of open science and transparency are quickly expanding in terms of access, reproducibility and replication initiatives worldwide. Nevertheless, so far the implications of openness are less discussed regarding a basic academic process: writing. In this presentation I analyze the limitations of traditional closed writing wokflows and introduce practices of academic writing in plain text editors in terms of replication and version control, as suggested by Healey (2018) in "The plain person's guide to plain text social science". Nevertheless, departing from Healey's suggestion of a writing workflow based on the editor Emacs, I propose an alternative workflow using the text editor Atom (also free and open source). Some of the Atom's advantages to be presented are citing directly from Zotero/Bibtex, integration of LaTeX, Markdown and R languages, and easy formatting from plain text to pdf, html and/or word via Pandoc. This constitutes a simple and friendly environment for open writing, which constitutes a baseline for introducing further transparent research practices.