Indicator-considerations eclipse other judgments on the shop-floor | Keynote Sarah de Rijcke ESA Prague, 26 August 2015

This invited lecture at the ESA conference in Prague drew on insights from the Leiden Manifesto and from two recent research projects at our institute in the Evaluation Practices in Context research group. These research projects show how indicators influence knowledge production in the life sciences and social sciences, and how in- and exclusion mechanisms get built into the scientific system through certain uses of evaluative metrics. Our findings point to a rather self-referential focus on metrics and a lack of space for responsible, relevant research in the scientific practices under study. On the basis of these findings I argued in the talk that we need an alternative moral discourse in research assessment, centered around the need to address growing inequalities in the science system. Secondly, the talk considered the most pertinent issues for the community of sociologists from the Leiden Manifesto for research metrics (Hicks, Wouters, Waltman, De Rijcke & Rafols, Nature, 23 April 2015).

http://www.slideshare.net/sarahderijcke/slideshelf

See also:

Rushforth & De Rijcke (2015). Accounting for Impact? The Journal Impact Factor and the making of biomedical research in the Netherlands. Minerva, 53(2), 117-139.

De Rijcke, S. & Rushforth, A.D. (2015). To intervene, or not to intervene, is that the question? On the role of scientometrics in research evaluation. Journal of the Association for Information Science and Technology, 66 (9), 1954-1958.

Hicks, D., Wouters, P.F., Rafols, I., De Rijcke, S. & Waltman, L. (2015). The Leiden Manifesto for Research Metrics. Nature, 23 April 2015.

Hammarfelt & De Rijcke (2015). Accountability in Context: Effects of research evaluation systems on publication practices, disciplinary norms, and individual working routines in the faculty of Arts at Uppsala University. Research Evaluation, 24(1), 63-77.

In Search of Excellence? Debating the Merits of Introducing an Elite Dutch University Model

Report by Alex Rushforth

Should the Netherlands strive for excellence in its university systems? Will maintaining quality suffice? This was the topic of a recent panel debate at the WTMC annual meeting on 21 November 2014 in De Balie, Amsterdam. Organised and chaired by Willem Halffman, the session focused on an article published by Barend Van Der Meulen in the national newspaper De Volkskrant, which advocated the need to produce two excellent universities which excel on internationally published rankings, thereby creating a new top-tier in the Dutch higher education system.

Both van der Meulen and Halffman presented their views, with an opposing position also coming from Sally Wyatt. Completing the panel, CWTS’s very own Paul Wouters provided results from recent empirical work about rankings.

Barend van der Meulen’s call for an elite university stemmed from the fact Dutch universities perennially sit outside of the top-50 in Shanghai and Times Higher Education rankings. For him the message is clear: the Netherlands is repeatedly failing to enhance its reputation as an elite player among global universities, a position which ought to cause concern. Van der Meulen stated that his call for an elite university model is part of a need to create an expanded repertoire of what universities are and what they should do in the Netherlands. The pursuit of rankings through this vehicle is therefore tightly coupled with a rejection of the status quo. Rankings are a social technology which ought to be harnessed for quality improvement and as tools through which to promote democratic participation by equipping students and policymakers with tools to make judgments and exert some influence over universities. Alternative modes of evaluation like peer review provide closed systems in which only other academics can make judgments, leaving university activities unaccountable to external modes of evaluation. This ‘ivory tower’ situation reminiscent of the 1980s is an image Van Der Meulen wishes to escape from, as ultimately it damages credibility and legitimacy of universities. The reliance on public money for research and education makes the moral case for university improvement and accountability particularly pressing in the Netherlands. For Van Der Meulen, the ‘good enough’ university (see Wyatt’s argument below) is not enough, given that excellence is imposing itself as a viable and increasingly important alternative.

First to oppose the motion in favour of elite universities was Willem Hallfman, whose talk built on a reply co-authored with Roland Bal, also in De Volkskrant. In the talk Halffman questioned the very foundations of the idea that ‘excellence’ ought to be pursued. Drawing unflattering comparisons between the research budget of Harvard University and that of the entire Netherlands, it was argued that competing within a global superleague would require a radical expansion of existing research budgets and wage structures across the Dutch university system, which he felt unrealistic and unreasonable against a backdrop of crisis in public finances. As well as reproducing national elites, Halffman also questioned the desirability of ranking systems which promote academic stars and the consequences this brings to institutions of science in general and Dutch universities in particular. Football-style league tables provide poor models on which to rate universities, as in contrast with sport where a winner-takes-all logic is central, for universities embodying a broad repertoire of societal functions, it is not clear what ‘winning’ means and how this would be made visible and commensurable through performance indicators.

Sally Wyatt recounted her personal experiences of the shock she encountered when studying and working in British universities in the 1980s, having grown-up in Canada within a period of prosperity and social mobility. These experiences fired a series of warning shots not to go down a road of pursuing excellence. When a move to the Netherlands came about in 1999, it promised her an oasis away from the turmoil the British university system had faced as a result of Thatcherite policy reforms. With the emergence of the Research Assessment Exercise (RAE) and its ranking logic comes also a rise in managerial positions and policies, decline in working conditions, and a widening gender gap. Gone also was a latent class system engrained in the culture of universities, with dominant elite institutions the site of social stratification reproduced across generations, which rankings merely encourage and reinforce. Despite erosion of certain positive attributes in universities since her arrival in the Netherlands, Wyatt argued that the Dutch system still preserves enough of a ‘level-playing field’ in terms of funding allocation to merit fierce resistance to any introduction of an elite university model. For Wyatt sometimes it is better to promote the ‘good enough’ than to chase an imperialist and elitist vision of ‘excellence’.

Drawing on work on university and hospital rankings carried-out with Sarah De Rijcke (CWTS), Iris Wallenburg and Roland Bal (Erasmus MC, Rotterdam), Paul Wouters’ talk advocated the need for a more fine-grained STS investigations into the kinds of work that goes into rankings, who is doing it, and in what situations. What is at stake in studying rankings then is not simply the critique of this or that tool, but a more pervasive (and sometimes invisible) logic/set of practices encountered across public organisations like universities and hospitals. Wouters advocated a move towards combining audit society critiques (which tend to be top-down) with STS insights into how ranking is practiced across various organisational levels in universities. This would provide a more promising platform through which to inform debates of the kind playing-out over the desirability of the elite university.

So the contrast between positions was stark. Are rankings – these seemingly ubiquitous ordering mechanisms of contemporary social life – something the Netherlands can afford to back away from in governing its universities? If they are being pursued anyway, shouldn’t policy intervene and assist a more systematic pursuit up the rankings which would enable more pronounced successes? Or is it necessary to oppose the very notion that the Netherlands needs to excel in a ‘globally competitive’ race, particularly given the seeming arbitrariness of many of the metrics according to which prestige gets attributed via ranking mechanisms? Despite polarization on what is to be done, potential for extending STS’s conceptual and empirical apparatus to mediate these discussions seemed to strike a chord among panelists and the audience alike. No doubt this stimulating debate touches on a set of issues that will not be going away quickly, and is one on which the WTMC community is surely well placed to intervene.

Ethics and misconduct – Review of a play organized by the Young Academy (KNAW)

This is a guest blog post by Joost Kosten. Joost is PhD student at CWTS and member of the EPIC working group. His research focuses on the use of research indicators from the perspective of public policy. Joost obtained an MSc in Public Administration (Leiden University) and was also trained in Political Science (Stockholm University) and Law (VU University Amsterdam).

Scientific (mis)conduct – The sins, the drama, the identification

On Tuesday November 18th 2014 the Young Academy of the Royal Netherlands Academy of Sciences organized a performance of the play Gewetenschap by Tony Maples at Leiden University. These weeks, Pandemonia Science Theater is on tour in the Netherlands to perform this piece at several universities. Gewetenschap was inspired by occasional troubles with respect to ethics and integrity which recently occurred in Dutch science and scholarship. Although these troubles concerned grave violations of the scientific code of conduct (i.e., the cardinal sins of fraud, fabrication, and plagiarism) the play focusses on common dilemma’s in a researcher’s everyday life. The title Gewetenschap is a non-existent word, which combines the Dutch words geweten (conscience) and wetenschap (science).

The playwright used confidential interviews with members of the Young Academy to gain insight into the most frequently occurring ethical dilemma’s researchers have to deal with. Professor Karin de Zwaan is a research group leader who has hardly any time to do research herself. She puts much effort in organizing grants, attracting new students and organizing her research group. Post-doc Jeroen Dreef is a very active researcher who does not have enough time to take organizational responsibilities serious. A tenure track is all he wants. Given their other important activities, Karin and Jeroen hardly have any time to supervise PhD student Lotte. One could question the type of support they do give her.

At times, given the reaction on scenes of the drama piece, the topics presented were clearly recognized by the audience. Afterwards, the dilemma’s touched upon during the play are presented by prof. Bas Haring. The audience discusses the following topics:

• Is there a conflict between the research topics a researcher likes himself and what the research group expects her/him to do?
• In one of the scenes, the researchers were delighted because of the acceptance of a publication. Haring asks if that exhibits “natural behaviour”. Shouldn’t a researcher be happy with good results instead of a publication being accepted? One of the participants replies that a publication functions as a reward.
• What do you do with your data? Is endless application of a diversity analysis methods until you find nice results a responsible approach?
• What about impact factors (IF)? Bas Haring himself says his IF is 0. “Do you think I am an idiot?” Which role do numbers such as the IF play in your opinion about colleagues? There seems to be quite a diversity of opinions. An early career research says everone knows these numbers are nonsense. An experienced scientist points out that there is a correlation between scores and quality. Someone else expresses his optimism since he expects that this focus on numbers will be over with ten years. This causes another to respond that in the past there was competition too, but in a different way.
• When is someone a co-author? This question results in a lively debate. Apparently, there are considerable differences from field to field. In the medical fields, a co-authorship can be a way to express gratitude to authors who have played a vital role in a research project, such as people who could organize experimental subjects. In this way, a co-authorship becomes a tradeable commodity. A medicine professor points out that in his field, co-authorships can be used to compare a curriculum vitae with the development of status as a researcher. Thus, it can be used as a criterion to judge grant proposals. A good researcher should start with first position co-authorships, later on should have co-authorships somewhere in between the first and last author, and should end his career with papers in which has co-authorships in the last position. Thus, the further the career has been developed, the more the name of the other should be in the final part of the author list. Another participant states that one can deal with co-authorships in three different ways: 1. Co-authors should always have full responsibility for everything in the paper. 2. Similar to openness which is given at the end of a movie, co-authors should clarify what each co-author’s contribution was. 3. Only those who really contributed in writing a paper can be a co-author. The participant admits that this last proposal works in his own field but might not work in other fields.
• Can a researcher exaggerate his findings if he presents them to journalists? Should you keep control over a journalist’s work in order to avoid that he will present things differently? Is it allowed to present untruth information in order to help support your case, just to avoid that a proper scientific argumentation will be too complex for the man in the street?
• Is it allowed to to present your work as having more societal relevance than you really expect? One of the reactions is that researchers are forced to express the societal relevance of their work when they apply for a grant. From the very nature of scientific research it is hardly possible to clearly indicate what society will gain from the results.
• What does a good relationship between a PhD-student and a supervisor look like? What is a good balance between serving the interests of PhD students, serving organizational interests (e.g. the future of the organization by attracting new students and grants), and the own interest of the researcher?

The discussion did not concentrate on the following dilemma´s presented in Gewetenschap:

• To what extent are requirements for grant proposals contradictory? On the one hand, researchers are expected to think ‘out-of-the-box’ while on the other hand they should meet a large amount of requirements. Moreover, should one propose new ideas including the risks which come along, or is it better to walk on the beaten path in order to guarantee successes?
• Should colleagues who did not show respect be served with the same sauce if you have a chance to review their work? Should you always judge scientific work on its merits? Are there any principles of ‘due process’ which should guide peer review?
• Whose are the data if someone contributed to them but moves to another research group or institute?

 

Developing guiding principles and standards in the field of evaluation – lessons learned

This is a guest blog post by professor Peter Dahler-Larsen. The reflections below are a follow-up of his keynote at the STI conference in Leiden (3-5 September 2014) and the special session at STI on the development of quality standards for science & technology indicators. Dahler-Larsen holds a chair at the Department of Political Science, University of Copenhagen. He is former president of the European Evaluation Society and author of The Evaluation Society (Stanford University Press, 2012).

Lessons learned about the development of guiding principles and standards in the field of evaluation – A personal reflection

Professor Peter Dahler-Larsen, 5 October 2014

Guidelines are symbolic, not regulatory

The limited institutional status of guiding principles and standards should be understood as a starting point for the debate. In the initial phases of development of such standards and guidelines, people often have very strong views. But only the state can enforce laws. To the extent that guidelines and standards merely express some official views of a professional association who has no institutional power to enforce them, standards and guidelines will have limited direct consequences for practitioners. The discussion becomes clearer once it is recognized that standards and guidelines thus primarily have a symbolic and communicative function, not a regulatory one. Practitioners will continue to be free to do whatever kind of practice they like, also after guidelines have been adopted.

Design a process of debate and involvement

All members of a professional association should have a possibility to comment on a draft version of guidelines/standards. An important component in the adoption of guidelines/standards is the design of a proper organizational process that involves the composition of a draft by a select group of recognized experts, an open debate among members, and an official procedure for the adoption of standards/guidelines as organizational policy.

Acknowledge the difference between minimum and maximum standards

Minimal standards must be complied with in all situations. Maximum standards are ideal principles worth striving for, although they will not be accomplished in any particular situation. It often turns out that there will be many maximum principles in a set of guidelines, although that is not what most people believe is “standards.” For that reason I personally prefer the term guidelines or guiding principles rather that “standards.”

Think carefully about guidelines and methodological pluralism

Advocates of a particular method often think that methodological rules connected to their own method defines quality as such in the whole field. For that reason, they are likely to insert their own methodological rules into the set of guidelines. As a consequence, guidelines can be used politically to promote one set of methods or one particular paradigm rather than another. Great care should be exercised in the formulation of guidelines to make sure that pluralism remains protected. For example, in evaluation the rule is that if you subscribe to a particular method, you should have high competence in the chosen method. But that goes for all methods.

Get beyond the “but that´s obvious” argument

Some argue that it is futile to formulate a set of guidelines because at that level of generality, it is only possible to state some very broad and obvious principles with which every sensible person must agree. The argument sounds plausible when you hear it, but my experience suggests otherwise for a number of reasons. First, some people have just not thought about a very bad practice (for example, doing evaluation without written Terms of Reference). Once you see, that someone has formulated a guideline against this, you are likely to start paying attention to the problem. Just because a principle is obvious to some, does not mean that it is obvious to all. Second, although there may be general agreement about a principle (such as “do no unnecessary harm” or “take general social welfare into account”), there can be strong disagreement about the interpretations and implications of the principle in practice.  Third, a good set of guiding principles will often comprise at least two principles that are somewhat in tension with each other, for example the principle of being quick and useful versus the principle of being scientifically rigorous. To sort out exactly which kind of tension between these two principles one can live with in a concrete case turns out to be a matter of complicated professional judgment. So, get beyond the “that´s obvious” argument.

Recognize the fruitful uses of guidelines

Among the most important uses of guidelines in evaluation are:

– In application situations, good evaluators can explain their practice with reference to broader principles

– In conferences, guidelines can stimulate insightful professional discussions about how to handle complicated cases

– Books and journals can make use of guidelines as inspiration for the development of an ethical awareness among practitioners. For example, google Michael Morris´ work in the field of evaluation.

– There is great use of guidelines in teaching and in other forms of socialization of evaluators.

Respect the multiplicity of organizations

If, say, the European Evaluation Society wants to adopt a set of guidelines, it should be respected that, say, the German and the Swiss association already have their own guidelines. Furthermore, some professional associations (say, psychologists) also have guidelines. A professional association should take such overlaps seriously and find ways to exchange views and experiences with guidelines across national and organizational borders.

Professionals are not alone, but relations can be described in guidelines, too

It is often debated that one of the major problems in bad evaluation practice is the behavior of commissioners. Some therefore think that guidelines describing good evaluation practice are in vain until the behavior of commissioners (and perhaps other users of evaluation) are included in the guidelines, too. However, there is no particular reason why the guidelines cannot describe a good relation and a good interaction between commissioners and evaluators. Remember, guidelines have no regulatory power. They express merely the official norms of the professional association. Evaluators are allowed to express what they think a good commissioner should do or not do. In fact, explicit guidelines can help clarify mutual and reciprocal role expectations.

Allow for regular reflection, evaluation and revision of guidelines

At regular intervals, guidelines should be debated, evaluated and revised. The AEA guidelines, for example, have been revised and now reflect values regarding culturally competent evaluation that was not in earlier versions. Guidelines are organic and reflect a particular socio-historical situation.

Sources:

Michael Morris (2008). Evaluation Ethics for Best Practice. Guilford Press.

American Evaluation Association Guiding principles

How does science go wrong?

We are happy to announce that our abstract got accepted for the 2014 Conference of the European Consortium for Political Research (ECPR), which will be held in Glasgow from 3-6 September. Our paper is selected for a panel on ‘The role of ideas and indicators in science policies and research management’, organised by Luis Sanz-Menéndez and Laura Cruz-Castro (both at CSIC-IPP).

Title of our paper: How does science go wrong?

“Science is in need of fundamental reform.” In 2013, five Dutch researchers took the lead in what they hope will become a strong movement for change in the governance of science and scholarship: Science in Transition. SiT appears to voice concerns heard beyond national borders about the need for change in the governance of science (cf. The Economist 19 October 2013; THE 23 Jan. 2014; Nature 16 Oct. 2013; Die Zeit 5 Jan. 2014). One of the most hotly debated concerns is quality control, and it encompasses the implications of a perceived increasing publication pressure, purported flaws in the peer review system, impact factor manipulation, irreproducibility of results, and the need for new forms of data quality management.

One could argue that SiT landed in fertile ground. In recent years, a number of severe fraud cases drew attention to possible ‘perverse effects’ in the management system of science and scholarship. Partly due to the juicy aspects of most cases of misconduct, these debates tend to focus on ‘bad apples’ and shy away from more fundamental problems in the governance of science and scholarship.

Our paper articulates how key actors construct the notion of ‘quality’ in these debates, and how they respond to each other’s position. By making these constructions explicit, we shift focus back to the self-reinforcing ‘performance loops’ that most researchers are caught up in at present. Our methodology is a combination of the mapping of the dynamics of media waves (Vasterman, 2005) and discourse analysis (Gilbert & Mulkay, 1984).

References

A revolutionary mission statement: improve the world. Times Higher Education, 23 January 2014.

Chalmers, I., Bracken, M. B., Djulbegovic, B., Garattini, S., Grant, J., Gülmezoglu, A. M., Oliver, S. (2014). How to increase value and reduce waste when research priorities are set. The Lancet, 383 (9912), 156–165.

Gilbert, G. N., & Mulkay, M. J. (1984). Opening Pandora’s Box. A Sociological Analysis of Scientists’ Discourse. Cambridge: Cambridge University Press.

Research evaluation: Impact. (2013). Nature, 502(7471), 287–287.

Rettet die Wissenschaft!: “Die Folgekosten können hoch sein.” Die Zeit, 5 January 2014.

Trouble at the lab. The Economist, 19 October 2013.

Vasterman, P. L. M. (2005). Media-Hype. European Journal of Communication , 20 (4 ), 508–530.