Course: Rules of good scientific practice
During the course, Henrik Hartmann will provide some background knowledge. Be prepared to get involved into a joint exploration good and bad scientific practice – and the grey area in between.
Since good scientific conduct is crucial for your own research activities and interactions between the scientific community and society at large all IMPRS members should deal with these issues and participate in one workshop like this.
January 21, 2014
Seminar room B0.002
We will start at 9 am and end in the early / mid afternoon.
register here by January 08, 2014.
Rules of good scientific practice – why and how to do honest research Good scientific practice is based on honesty and should promote public trust in scientific findings. While it can be assumed that most scientists are honest by intention, there are several sources of scientific misconduct. Lack of care applied to the scientific method but also deliberate fraud are examples of violation of good scientific practices. These violations, whether intentional or not, can spoil the public’s perception of scientific research and can cause major damages to science and – even more important – to society.
This course builds on the Max Planck Society’s ‘Rules of good scientific practice’ ( --> Good Practice <--), the ‘Rules of procedure in cases of suspected scientific misconduct’ ( --> Misconduct <--) as well as the 'Guidelines and rules on a responsible approach to freedom of research and research risks' ( --> Freedom & Risks <--). During the lecture, we will discover the Society’s ruling of good scientific practices, consider the ethical aspects of research and learn how to proceed in cases of suspected scientific misconduct.
Furthermore, the course will elaborate several aspects of scientific tasks that are particularly important to early-career scientists. Researchers will have insights into the peer reviewing process and will be shown how to do a good peer review. They will discover the dangers of plagiarism and learn how to avoid these by applying adequate acknowledgement and citation. Moreover, we will examine the structure of manuscripts and show what information must be given in each of its sections. General editing rules for manuscript improvement will also be part of the course. Last but not least, we will have a quick look at statistical issues by highlighting the limitations of traditional null hypothesis testing (NHT) and give details on one of several alternatives to NHT, namely information-theory based multi-model inference.
The course is a project-based application of the rules of good scientific conduct. Researchers will learn how to avoid potential pitfalls in scientific conduct by understanding the psychological mechanisms that could predispose them to misconduct. Based on a real-world example (see below), researchers will analytically describe the history of events. What happened and why? What rules have been infringed? PhD researchers will develop a written conceptual essay on how these misconducts could have been prevented. The essay should draw on (but may not necessarily be limited to) a selection of literature (see below) which encompasses several aspects of scientific conduct. Among these are psychological and ethical aspects, but also technical issues like data acquisition, storage and processing as well as statistical, publication and media issues.
- Form an idea of what is good and bad scientific practice.
- Think about how to differentiate between the two.
- Bring an example of bad scientific conduct to the seminar.
- Write an analysis of a known case of scientific misconduct: either based on the suggestion that Henrik presents during the seminar or a topic of your choice (best discuss this with Henrik first).
Participants will be evaluated and credited according to the effort they put into their essay. Evaluation criteria are:
- Have all events of misconducts as well as their source (e.g., conflict of interest, lack of openness to criticism) been properly identified?
- Are the suggestions for avoiding misconducts based on a thorough understanding of the related concepts/literature?
- Are the suggestions conclusive?
- Was information/literature used beyond the one suggested?
- Have new ideas been brought into discussion?
- Formal presentation of the essay (content, structure, linguistic quality, logical flow etc.).
Participants will have a three-week period for writing the essay (5-10 pages). Further information will be given during the lecture. The credits you can obtain for participating in this seminar range from 0.2 for mere attendance to 1.0 for submission of a good essay.
Our working example is the MMR vaccine (measles, mumps, rubella) scandal that has been triggered off in 1998 by an observational study of children suffering from some sort of autistic behavior. Because most of these children had received a MMR vaccination shortly before the onset of behavioral symptoms, the authors suggested (and pathologically explain) a possible link between MMR and autism. The findings of the study were heavily exploited by the media and led to a major health crisis in the UK. Over the past 12 years, the main author and its research were under tight scrutiny by an investigative journalist and the British General Medical Council. Their findings reveal one of the biggest scandals in medical research...
- Wakefield et al. 1998_ MMR link to autism.pdf
- Chen & DeStefano 1998_Vaccine adverse events causal coincidental.pdf
The MMR scandal - history and consequences
- Deer online material_Nailed: Wakefield and the MMR autism fraud.pdf
- Godlee et al. 2011_Article linking MMR vaccine autism fraudulent .pdf
- Deer 2011_How the vaccine crisis was meant to make money.pdf
Science in trouble - scientists behaving badly and the role of media
- Claxton 2005a_Scientific authorship window into fraud.pdf
- Claxton 2005b_Scientific authorship history issues practices guidelines.pdf
- Martinson et al. 2005_Scientists behaving badly.pdf
- Franzen et al. 2007_Fraud causes culprits perceived by science media.pdf
Psychological and ethical aspects of science
- Glass 1965_Ethical basis science.pdf
- Loehle 1987_Hypothesis testing ecology.pdf
- Gross 2009_Broken trust lessons from vaccine autism war.pdf
- see also this page --> Ethics in science <-- for literature on general ethical aspects of science
Peer review process
* Relman 1990_Peer review scientific journals what good is it.pdf
Writing scientific articles
* Schulmann 1996_How to write a scientific paper_FUNNY.pdf
- Carpenter 2001_How to Write a Scientific Paper.pdf
- Sand-Jensen 2007_How to write consitently boring papers.pdf
Methodological and statistical issues
- Burnham & Anderson 2002_Data dredging.pdf In: Burnham & Anderson (2002) Model selection and multimodel inference: a practical information theoretic approach. Springer. 455 p. --> Available at the BGC/ICE library: Bar code J0014499, location STA40/072
- Kallo 2010_Data dredging.pdf
- Hurlbert 1984_Pseudoreplication design ecological field experiments.pdf
- Anderson et al. 2000_Null hypothesis testing problems prevelance alternative.pdf
- Anderson et al. 2001_Concerns finding spurious effects.pdf
- Stephens et al. 2005_Information theory hypothesis testing call pluralism.pdf
- Lukas et al. 2007_Concerns regarding call pluralism Stephens et al. 2005.pdf
- Johnson & Omland 2004_Model selection ecology evolution.pdf
- Stephens et al. 2006_Inference ecology evolution.pdf
7 out of 11 participants filled in the survey by February 19, 2014. Thanks a lot for participating in this survey! Your feedback is valuable because it helps the instructors and organizers to improve the individual modules and the general structure of the course.
The survey results are available here. Statistics and statements should not be taken as an exhaustive or exclusive list.