Category Archives: incentives

How to review a paper

Including a Checklist for Hypothesis Testing Research Reports *

See https://osf.io/6cw7b/ for a pdf of this post

 

Academia critically relies on our efforts as peer reviewers to evaluate the quality of research that is published in journals. Reading the reviews of others, I have noticed that the quality varies considerably, and that some reviews are not helpful. The added value of a journal article above and beyond the original manuscript or a non-reviewed preprint is in the changes the authors made in response to the reviews. Through our reviews, we can help to improve the quality of the research. This memo provides guidance on how to review a paper, partly inspired by suggestions provided by Alexander (2005), Lee (1995) and the Committee on Publication Ethics (2017). To improve the quality of the peer review process, I suggest that you use the following guidelines. Some of the guidelines – particularly the criteria at the end of this post – are peculiar for the kind of research that I tend to review – hypothesis testing research reports relying on administrative data and surveys, sometimes with an experimental design. But let me start with guidelines that I believe make sense for all research.

Things to check before you accept the invitation
First, I encourage you to check whether the journal aligns with your vision of science. I find that a journal published by an exploitative publisher making a profit in the range of 30%-40% is not worth my time. A journal that I have submitted my own work to and gave me good reviews is worth the number of reviews I received for my article. The review of a revised version of the paper does not count as a separate paper.
Next, I check whether I am the right person to review the paper. I think it is a good principle to describe my disciplinary background and expertise in relation to the manuscript I am invited to review. Reviewers do not need to be experts in all respects. If you do not have useful expertise to improve the paper, politely decline.

Then I check whether I know the author(s). If I do, and I have not collaborated with the author(s), if I am not currently collaborating or planning to do so, I describe how I know the author(s) and ask the editor whether it is appropriate for me to review the paper. If I have a conflict of interest, I notify the editor and politely decline. It is a good principle to let the editor know immediately if you are unable to review a paper, so the editor can start to look for someone else to review the paper. Your non-response means a delay for the authors and the editor.

Sometimes I get requests to review a paper that I have reviewed before, for a conference or another journal. In these cases I let the editor know and ask the editor whether she would like to see the previous review. For the editor it will be useful to know whether the current manuscript is the same as the version, or includes revisions.

Finally, I check whether the authors have made the data and code available. I have made it a requirement that authors have to fulfil before I accept an invitation to review their work. An exception can be made for data that would be illegal or dangerous to make available, such as datasets that contain identifying information that cannot be removed. In most cases, however, the authors can provide at least partial access to the data by excluding variables that contain personal information.

A paper that does not provide access to the data analyzed and the code used to produce the results in the paper is not worth my time. If the paper does not provide a link to the data and the analysis script, I ask the editor to ask the authors to provide the data and the code. I encourage you to do the same. Almost always the editor is willing to ask the authors to provide access. If the editor does not respond to your request, that is a red flag to me. I decline future invitation requests from the journal. If the authors do not respond to the editor’s request, or are unwilling to provide access to the data and code, that is a red flag for the editor.

The tone of the review
When I write a review, I think of the ‘golden rule’: treat others as you would like to be treated. I write the review report that I would have liked to receive if I had been the author. I use the following principles:

  • Be honest but constructive. You are not at war. There is no need to burn a paper to the ground.
  • Avoid addressing the authors personally. Say: “the paper could benefit from…” instead of “the authors need”.
  • Stay close to the facts. Do not speculate about reasons why the authors have made certain choices beyond the arguments stated in the paper.
  • Take a developmental approach. Any paper will contain flaws and imperfections. Your job is to improve science by identifying problems and suggesting ways to repair them. Think with the authors about ways they can improve the paper in such a way that it benefits collective scholarship. After a quick glance at the paper, I determine whether I think the paper has the potential to be published, perhaps after revisions. If I think the paper is beyond repair, I explain this to the editor.
  • Try to see beyond bad writing style and mistakes in spelling. Also be mindful of disciplinary and cultural differences between the authors and yourself.

The substance of the advice
In my view, it is a good principle to begin the review report by describing your expertise and the way you reviewed the paper. If you searched for literature, checked the data and verified the results, ran additional analyses, state this. It will allow the editor to adjudicate the review.

Then give a brief overview of the paper. If the invitation asks you to provide a general recommendation, consider whether you’d like to give one. Typically, you are invited to recommend ‘reject’, ‘revise & resubmit’ – with major or minor revisions, or ‘accept’. Because the recommendation is the first thing the editor wants to know it is convenient to state it early in the review.

When giving such a recommendation, I start from the assumption that the authors have invested a great deal of time in the paper and that they want to improve it. Also I consider the desk-rejection rate at the journal. If the editor sent the paper out for review, she probably thinks it has the potential to be published.

To get to the general recommendation, I list the strengths and the weaknesses of the paper. To ease the message you can use the sandwich principle: start with the strengths, then discuss the weaknesses, and conclude with an encouragement.

For authors and editors alike it is convenient to give actionable advice. For the weaknesses in the paper I suggest ways to repair them. I distinguish major issues such as not discussing alternative explanations from minor issues such as missing references and typos. It is convenient for both the editor and the authors to number your suggestions.

The strengths could be points that the authors are underselling. In that case, I identify them as strengths that the authors can emphasize more strongly.

It is handy to refer to issues with direct quotes and page numbers. To refer to the previous sentence: “As the paper states on page 3, [use] “direct quotes and page numbers””.

In 2016 I have started to sign my reviews. This is an accountability device: by exposing who I am to the authors of the paper I’m reviewing, I set higher standards for myself. I encourage you to think about this as an option, though I can imagine that you may not want to risk retribution as a graduate student or an early career researcher. Also some editors do not appreciate signed reviews and may take away your identifying information.

How to organize the review work
Usually, I read a paper twice. First, I go over the paper superficially and quickly. I do not read it closely. This gets me a sense of where the authors are going. After the first superficial reading, I determine whether the paper is good enough to be revised and resubmitted, and if so, I provide more detailed comments. After the report is done, I revisit my initial recommendation.

The second time I go over the paper, I do a very close reading. Because the authors had a word limit, I assume that literally every word in the manuscript is absolutely necessary – the paper should have no repetitions. Some of the information may be in the supplementary information provided with the paper.

Below you find a checklist of things I look for in a paper. The checklist reflects the kind of research that I tend to review, which is typically testing a set of hypotheses based on theory and previous research with data from surveys, experiments, or archival sources. For other types of research – such as non-empirical papers, exploratory reports, and studies based on interviews or ethnographic material – the checklist is less appropriate. The checklist may also be helpful for authors preparing research reports.

I realize that this is an extensive set of criteria for reviews. It sets the bar pretty high. A review checking each of the criteria will take you at least three hours, but more likely between five and eight hours. As a reviewer, I do not always check all criteria myself. Some of the criteria do not necessarily have to be done by peer reviewers. For instance, some journals employ data editors who check whether data and code provided by authors produce the results reported.

I do hope that journals and editors can get to a consensus on a set of minimum criteria that the peer review process should cover, or at least provide clarity about the criteria that they do check.

After the review
If the authors have revised their paper, it is a good principle to avoid making new demands for the second round that you have not made before. Otherwise the revise and resubmit path can be very long.

 

References
Alexander, G.R. (2005). A Guide to Reviewing Manuscripts. Maternal and Child Health Journal, 9 (1): 113-117. https://doi.org/10.1007/s10995-005-2423-y
Committee on Publication Ethics Council (2017). Ethical guidelines for peer reviewers. https://publicationethics.org/files/Ethical_Guidelines_For_Peer_Reviewers_2.pdf
Lee, A.S. (1995). Reviewing a manuscript for publication. Journal of Operations Management, 13: 87-92. https://doi.org/10.1016/0272-6963(95)94762-W

 

Review checklist for hypothesis testing reports

Research question

  1. Is it clear from the beginning what the research question is? If it is in the title, that’s good. In the first part of the abstract is good too. Is it at the end of the introduction section? In most cases that is too late.
  2. Is it clearly formulated? By the research question alone, can you tell what the paper is about?
  3. Does the research question align with what the paper actually does – or can do – to answer it?
  4. Is it important to know the answer to the research question for previous theory and methods?
  5. Does the paper address a question that is important from a societal or practical point of view?

 

Research design

  1. Does the research design align with the research question? If the question is descriptive, do the data actually allow for a representative and valid description? If the question is a causal question, do the data allow for causal inference? If not, ask the authors to report ‘associations’ rather than ‘effects’.
  2. Is the research design clearly described? Does the paper report all the steps taken to collect the data?
  3. Does the paper identify mediators of the alleged effect? Does the paper identify moderators as boundary conditions?
  4. Is the research design waterproof? Does the study allow for alternative interpretations?
  5. Has the research design been preregistered? Does the paper refer to a public URL where the preregistration is posted? Does the preregistration include a statistical power analysis? Is the number of observations sufficient for statistical tests of hypotheses? Are deviations from the preregistered design reported?
  6. Has the experiment been approved by an Internal or Ethics Review Board (IRB/ERB)? What is the IRB registration number?

 

Theory

  1. Does the paper identify multiple relevant theories?
  2. Does the theory section specify hypotheses? Have the hypotheses been formulated before the data were collected? Before the data were analyzed?
  3. Do hypotheses specify arguments why two variables are associated? Have alternative arguments been considered?
  4. Is the literature review complete? Does the paper cover the most relevant previous studies, also outside the discipline? Provide references to research that is not covered in the paper, but should definitely be cited.

 

Data & Methods

  1. Target group – Is it identified? If mankind, is the sample a good sample of mankind? Does it cover all relevant units?
  2. Sample – Does the paper identify the procedure used to obtain the sample from the target group? Is the sample a random sample? If not, has selective non-response been dealt with, examined, and have constraints on generality been identified as a limitation?
  3. Number of observations – What is the statistical power of the analysis? Does the paper report a power analysis?
  4. Measures – Does the paper provide the complete topic list, questionnaire, instructions for participants? To what extent are the measures used valid? Reliable?
  5. Descriptive statistics – Does the paper provide a table of descriptive statistics (minimum, maximum, mean, standard deviation, number of observations) for all variables in the analyses? If not, ask for such a table.
  6. Outliers – Does the paper identify treatment of outliers, if any?
  7. Is the multi-level structure (e.g., persons in time and space) identified and taken into account in an appropriate manner in the analysis? Are standard errors clustered?
  8. Does the paper report statistical mediation analyses for all hypothesized explanation(s)? Do the mediation analyses evaluate multiple pathways, or just one?
  9. Do the data allow for testing additional explanations that are not reported in the paper?

 

Results

  1. Can the results be reproduced from the data and code provided by the authors?
  2. Are the results robust to different specifications?

Conclusion

  1. Does the paper give a clear answer to the research question posed in the introduction?
  2. Does the paper identify implications for the theories tested, and are they justified?
  3. Does the paper identify implications for practice, and are they justified given the evidence presented?

 

Discussion

  1. Does the paper revisit the limitations of the data and methods?
  2. Does the paper suggest future research to repair the limitations?

 

Meta

  1. Does the paper have an author contribution note? Is it clear who did what?
  2. Are all analyses reported, if they are not in the main text, are they available in an online appendix?
  3. Are references up to date? Does the reference list include a reference to the dataset analyzed, including an URL/DOI?

 

 

* This work is licensed under a Creative Commons Attribution 4.0 International License. Thanks to colleagues at the Center for Philanthropic Studies at Vrije Universiteit Amsterdam, in particular Pamala Wiepking, Arjen de Wit, Theo Schuyt and Claire van Teunenbroek, for insightful comments on the first version. Thanks to Robin Banks, Pat Danahey Janin, Rense Corten, David Reinstein, Eleanor Brilliant, Claire Routley, Margaret Harris, Brenda Bushouse, Craig Furneaux, Angela Eikenberry, Jennifer Dodge, and Tracey Coule for responses to the second draft. The current text is the fourth draft. The most recent version of this paper is available as a preprint at https://doi.org/10.31219/osf.io/7ug4w. Suggestions continue to be welcome at r.bekkers@vu.nl.

Leave a comment

Filed under academic misconduct, data, experiments, fraud, helping, incentives, law, open science, sociology, survey research

Cut the crap, fund the research

We all spend way too much time preparing applications for research grants. This is a collective waste of time. For the 2019 vici grant scheme of the Netherlands Organization for Scientific Research (NWO) in which I recently participated, 87% of all applicants received no grant. Based on my own experiences, I made a conservative calculation (here is the excel file so you can check it yourself) of the total costs for all people involved. The costs total €18.7 million. Imagine how much research time that is worth!

Cost

Applicants account for the bulk of the costs. Taken together, all applicants invested €15.8 million euro in the grant competition. As an applicant, I read the call for proposals, first considered whether or not I would apply, decided yes, I read the guidelines for applications, discussed ideas with colleagues, read the literature, wrote a short draft of the proposal to invite research partners, then wrote the proposal text, formatted the application according to the guidelines, prepared a budget for approval, collected some new data and analyzed it, considered whether ethics review was necessary, created a data management plan, corresponded with: grants advisors, a budget controller, HR advisors, internal reviewers, my head of department, the dean, a coach, and with societal partners. I revised the application, revised the budget, and submitted the preproposal. I waited. And waited. Then I read the preproposal evaluation by the committee members, and wrote responses to the preproposal evaluation. I revised my draft application again, and submitted the full application. I waited. And waited. I read the external reviews, wrote responses to their comments, and submitted a rebuttal. I waited. And waited. Then I prepared a 5 minutes pitch for the interview by the committee, responded to questions, and waited. Imagine I would have spent all that time on actual research. Each applicant could have spent 971 hours on research instead.

Also the university support system spends a lot of resources preparing budgets, internal reviews, and training of candidates. I involved research partners and societal partners to support the proposal. I feel bad for wasting their time as well.

The procedure also puts a burden on external reviewers. At a conference I attended, one of the reviewers of my application identified herself and asked me what had happened with the review she had provided. She had not heard back from the grant agency. I told her that she was not the only one who had given an A+ evaluation, but that NWO had overruled it in its procedures.

For the entire vici competition, an amount of €46.5 million was available, for 32 grants to be awarded. The amount wasted is 40% of that amount! That is unacceptable.

It is time to stop wasting our time.

 

Note: In a previous version of this post, I assumed that the number of applicants was 100. This estimate was much too low. The grant competition website says that across all domains 242 proposals were submitted. I revised the cost calculation (v2) to reflect the actual number of applicants. Note that this calculation leaves out hours spent by researchers who eventually decided not to submit a (pre-)proposal. The calculation further assumes that 180 full proposals were submitted and 105 candidates were interviewed.

Update, February 26: In the previous the cost of the procedure for NWO was severely underestimated. According to the annual report of NWO, the total salary costs for its staff that handles grant applications is €72 million per year. In the revised cost calculation, I’m assuming staff spend 218 hours for the entire vici competition. This amount consists of €198k variable costs (checking applications, inviting reviewers, composing decision letters, informing applicants, informing reviewers, handling appeals by 10% of full proposals, and handling ‘WOB verzoeken’ = Freedom Of Information Act requests) and €20k fixed costs: preparing the call for proposals, organizing committee meetings to discuss applications and their evaluations, attending committee meetings, reporting on committee meetings, evaluating the procedure).

Leave a comment

Filed under academic misconduct, economics, incentives, policy evaluation, taxes, time use

Global Giving: Open Grant Proposal

Here’s an unusual thing for you to read: I am posting a brief description of a grant proposal that I will submit for the ‘vici’-competition of the Netherlands Organization for Scientific Research 2019 later this year. You can download the “pre-proposal” here. It is called “Global Giving”. With the study I aim to describe and explain philanthropy in a large number of countries across the world. I invite you to review the “pre-proposal” and suggest improvements; please use the comments box below, or write to me directly.

You may have heard the story that university researchers these days spend a lot of their time writing grant proposals for funding competitions. Also you may have heard the story that chances of success in such competitions are getting smaller and smaller. These stories are all true. But the story you seldom hear is how such competitions actually work: they are a source of stress, frustration, burnouts and depression, and a complete waste of the precious time of the smartest people in the world. Recently, Gross and Bergstrom found that “the effort researchers waste in writing proposals may be comparable to the total scientific value of the research that the funding supports”.

Remember the last time you saw the announcement of prize winners in a research grant competition? I have not heard a single voice in the choir of the many near-winners speak up: “Hey, I did not get a grant!” It is almost as if everybody wins all the time. It is not common in academia to be open about failures to win. How many vitaes you have seen recently contain a list of failures? This is a grave distortion of reality. Less than one in ten applications is succesful. This means that for each winning proposal there are at least nine proposals that did not get funding. I want you to know how much time is wasted by this procedure. So here I will be sharing my experiences with the upcoming ‘vici’-competition.

single-shot-santa

First let me tell you about the funny name of the competition. The name ‘vici’ derives from roman emperor Caesar’s famous phrase in Latin: ‘veni, vidi, vici’, which he allegedly used to describe a swift victory. The translation is: “I came, I saw, I conquered”. The Netherlands Organization for Scientific Research (‘Nederlandse organisatie voor Wetenschappelijk Onderzoek’, NWO) thought it fitting to use these names as titles of their personal grant schemes. The so-called ‘talent schemes’ are very much about the personal qualities of the applicant. The scheme heralds heroes. The fascination with talent goes against the very nature of science, where the value of an idea, method or result is not measured by the personality of the author, but by its validity and reliability. That is why peer review is often double blind and evaluators do not know who wrote the research report or proposal.

plt132

Yet in the talent scheme, the personality of the applicant is very important. The fascination with talent creates Matthew effects, first described in 1968 by Robert K. Merton. The name ‘Matthew effect’ derives from the biblical phrase “For to him who has will more be given” (Mark 4:25). Simply stated: success breeds success. Recently, this effect has been documented in the talent scheme by Thijs Bol, Matthijs de Vaan and Arnout van de Rijt. When two applicants are equally good but one – by mere chance – receives a grant and the other does not, the ‘winner’ is ascribed with talent and the ‘loser’ is not. The ‘winner’ then gets a tremendously higher chance of receiving future grants.

As a member of committees for the ‘veni’ competition I have seen how this works in practice. Applicants received scores for the quality of their proposal from expert reviewers before we interviewed them. When we had minimal differences between the expert reviewer scores of candidates – differing only in the second decimal – personal characteristics of the researchers such as their self-confidence and manner of speaking during the interview often made the difference between ‘winners’ and ‘losers’. Ultimately, such minute differences add up to dramatically higher chances to be a full professor 10 years later, as the analysis in Figure 4 of the Bol, De Vaan & Van de Rijt paper shows.

matthew

My career is in this graph. In 2005, I won a ‘veni’-grant, the early career grant that the Figure above is about. The grant gave me a lot of freedom for research and I enjoyed it tremendously. I am pretty certain that the freedom that the grant gave me paved the way for the full professorship that I was recently awarded, thirteen years later. But back then, the size of the grant did not feel right. I felt sorry for those who did not make it. I knew I was privileged, and the research money I obtained was more than I needed. It would be much better to reduce the size of grants, so that a larger number of researchers can be funded. Yet the scheme is there, and it is a rare opportunity for researchers in the Netherlands to get funding for their own ideas.

This is my third and final application for a vici-grant. The rules for submission of proposals in this competition limit the number of attempts to three. Why am I going public with this final attempt?

The Open Science Revolution

You will have heard about open science. Most likely you will associate it with the struggle to publish research articles without paywalls, the exploitation of government funded scientists by commercial publishers, and perhaps even with Plan S. You may also associate open science with the struggle to get researchers to publish the data and the code they used to get to their results. Perhaps you have heard about open peer review of research publications. But most likely you will not have heard about open grant review. This is because it rarely happens. I am not the first to publish my proposal; the Open Grants repository currently contains 160 grant proposals. These proposals were shared after the competitions had run. The RIO Journal published 52 grant proposals. This is only a fraction of all grant proposals being created, submitted and reviewed. The many advantages of open science are not limited to funded research, they also apply to research ideas and proposals. By publishing my grant proposal before the competition, the expert reviews, the recommendations of the committee, my responses and experiences with the review process, I am opening up the procedure of grant review as much as possible.

Stages in the NWO Talent Scheme Grant Review Procedure

Each round of this competition takes almost a year, and proceeds in eight stages:

  1. Pre-application – March 26, 2019 <– this is where we are now
  2. Non-binding advice from committee: submit full proposal, or not – Summer 2019
  3. Full proposal – end of August 2019
  4. Expert reviews – October 2019
  5. Rebuttal to criticism in expert reviews – end of October 2019
  6. Selection for interview – November 2019
  7. Interview – January or February 2020
  8. Grant, or not – March 2020

If you’re curious to learn how this application procedure works in practice,
check back in a few weeks. Your comments and suggestions on the ideas above and the pre-proposal are most welcome!

Leave a comment

Filed under altruism, charitable organizations, data, economics, empathy, experiments, fundraising, happiness, helping, household giving, incentives, methodology, open science, organ donation, philanthropy, politics, principle of care, psychology, regression analysis, regulation, sociology, statistical analysis, survey research, taxes, trends, trust, volunteering, wealth

Closing the Age of Competitive Science

In the prehistoric era of competitive science, researchers were like magicians: they earned a reputation for tricks that nobody could repeat and shared their secrets only with trusted disciples. In the new age of open science, researchers share by default, not only with peer reviewers and fellow researchers, but with the public at large. The transparency of open science reduces the temptation of private profit maximization and the collective inefficiency in information asymmetries inherent in competitive markets. In a seminar organized by the University Library at Vrije Universiteit Amsterdam on November 1, 2018, I discussed recent developments in open science and its implications for research careers and progress in knowledge discovery. The slides are posted here. The podcast is here.

2 Comments

Filed under academic misconduct, data, experiments, fraud, incentives, law, Netherlands, open science, statistical analysis, survey research, VU University

How not to solve the research competition crisis

Scientists across the globe spend a substantial part of their time writing research proposals for competitive grant schemes. Usually, less than one in seven proposals gets funded. Moreover, the level of competition and the waste of time invested in research proposals that do not receive funding are increasing.

The most important funder of science in the Netherlands, the Netherlands Organization for Scientific Research (NWO), is painfully aware of the research competition crisis. On April 4, 2017, more than one hundred of the nation’s scientists gathered in a conference to come up with solutions for the crisis. I was one of them.

The conference made clear that the key problem is that we have too many good candidates and high quality research proposals that cannot be funded with the current budget. Without an increase in the budget for research funding, however, that problem is unlikely to go away.

pipe-line-icon

Stan Gielen, the new director of NWO, opened the conference. Because the universities and NWO lack bargaining power in the government that determines the budget for NWO, he asked the scientists at the conference to think about ‘streamlining procedures’. In roundtable discussions, researchers talked about questions like: “How can the time it takes between a final ranking in a grant competition and the announcement of the result to applicants be reduced?”

Many proposals came up during the meeting. The more radical proposals were to discontinue funding for NWO altogether and to reallocate funding back to the universities, to give a larger number of smaller grants, to allocate funding through lotteries among top-rated applications, and the idea by Scheffer to give researchers voting rights on funding allocations. I left the meeting with an increased sense of urgency but with little hope for a solution. Gielen concluded the meeting with the promise to initiate conversations with the ministry for Education, Culture and Science about the results of the conference and to report back within six months.

Yesterday, NWO presented its proposals. None of the ideas above made it. Instead, a set of measures were announced that are unlikely to increase chances of funding. The press release does not say why ineffective measures were favored over effective measures.

Two of the proposals by NWO shift work to the universities, giving them responsibility in pre-evaluations of proposals. At the Vrije Universiteit Amsterdam we already make quite an investment in such pre-evaluations, but not all universities do so. Also the universities are now told to use an instrument to reduce the number of proposals: the financial guarantee. Also this proposal is akin to a measure we already had in place, the obligatory budget check. The financial guarantee is an additional hurdle applicants have to take.

The proposal to give non-funded but top-rated ERC proposals a second chance at NWO reduces some of the work for applicants, but does not increase chances for funding.

A final proposal is to ask applicants to work together with other applicants with related ideas. It may be a good idea for other reasons, but does not increase chances for funding.

 

Now what?

One of the causes of the problem that funding chances are declining is the reward that universities get for graduations of PhD candidates (‘promotiepremie’). This reward keeps up the supply of good researchers. PhD candidates are prepared and motivated for careers in science. But these careers are increasingly hard to get into. As long as the dissertation defense reward is in place, one long term solution is to change the curriculum in graduate schools, orienting them to non-academic careers.

Another long-term solution is to diversify funding sources for science. In the previous cabinets, the ministry of Economic Affairs has co-controlled funding allocations to what were labeled ‘topsectors’. Evaluations of this policy have been predominantly negative. One of the problems is that the total budget for science was not increased, but the available budget was partly reallocated for applied research in energy, water, logistics etcetera. It is unclear how the new government thinks about this, but it seems a safe bet not to have much hope for creative ideas from this side. But there is hope for a private sector solution.

There is a huge amount of wealth in the Netherlands that investment bankers are trying to invest responsibly. As a result of increases in wealth, the number of private foundations established that support research and innovation has increased strongly in the past two decades. These foundations are experimenting with new financial instruments like impact investing and venture philanthropy. The current infrastructure and education at universities, however, is totally unfit to tap into this potential of wealth. Which graduate program offers a course in creating a business case for investments in research?

 

 

 

1 Comment

Filed under Europe, incentives, policy evaluation, politics, VU University

Tools for the Evaluation of the Quality of Experimental Research

pdf of this post

Experiments can have important advantages above other research designs. The most important advantage of experiments concerns internal validity. Random assignment to treatment reduces the attribution problem and increases the possibilities for causal inference. An additional advantage is that control over participants reduces heterogeneity of treatment effects observed.

The extent to which these advantages are realized in the data depends on the design and execution of the experiment. Experiments have a higher quality if the sample size is larger, the theoretical concepts are more reliably measured, and have a higher validity. The sufficiency of the sample size can be checked with a power analysis. For most effect sizes in the social sciences, which are small (d = 0.2), a sample of 1300 participants is required to detect it at conventional significance levels (p < .05) and 95% power (see appendix). Also for a stronger effect size (0.4) more than 300 participants are required. The reliability of normative scale measures can be judged with Cronbach’s alpha. A rule of thumb for unidimensional scales is that alpha should be at least .63 for a scale consisting of 4 items, .68 for 5 items, .72 for 6 items, .75 for 7 items, and so on. The validity of measures should be justified theoretically and can be checked with a manipulation check, which should reveal a sizeable and significant association with the treatment variables.

The advantages of experiments are reduced if assignment to treatment is non-random and treatment effects are confounded. In addition, a variety of other problems may endanger internal validity. Shadish, Cook & Campbell (2002) provide a useful list of such problems.

Also it should be noted that experiments can have important disadvantages. The most important disadvantage is that the external validity of the findings is limited to the participants in the setting in which their behavior was observed. This disadvantage can be avoided by creating more realistic decision situations, for instance in natural field experiments, and by recruiting (non-‘WEIRD’) samples of participants that are more representative of the target population. As Henrich, Heine & Norenzayan (2010) noted, results based on samples of participants in Western, Educated, Industrialized, Rich and Democratic (WEIRD) countries have limited validity in the discovery of universal laws of human cognition, emotion or behavior.

Recently, experimental research paradigms have received fierce criticism. Results of research often cannot be reproduced (Open Science Collaboration, 2015), publication bias is ubiquitous (Ioannidis, 2005). It has become clear that there is a lot of undisclosed flexibility, in all phases of the empirical cycle. While these problems have been discussed widely in communities of researchers conducting experiments, they are by no means limited to one particular methodology or mode of data collection. It is likely that they also occur in communities of researchers using survey or interview data.

In the positivist paradigm that dominates experimental research, the empirical cycle starts with the formulation of a research question. To answer the question, hypotheses are formulated based on established theories and previous research findings. Then the research is designed, data are collected, a predetermined analysis plan is executed, results are interpreted, the research report is written and submitted for peer review. After the usual round(s) of revisions, the findings are incorporated in the body of knowledge.

The validity and reliability of results from experiments can be compromised in two ways. The first is by juggling with the order of phases in the empirical cycle. Researchers can decide to amend their research questions and hypotheses after they have seen the results of their analyses. Kerr (1989) labeled the practice of reformulating hypotheses HARKING: Hypothesizing After Results are Known. Amending hypotheses is not a problem when the goal of the research is to develop theories to be tested later, as in grounded theory or exploratory analyses (e.g., data mining). But in hypothesis-testing research harking is a problem, because it increases the likelihood of publishing false positives. Chance findings are interpreted post hoc as confirmations of hypotheses that a priori  are rather unlikely to be true. When these findings are published, they are unlikely to be reproducible by other researchers, creating research waste, and worse, reducing the reliability of published knowledge.

The second way the validity and reliability of results from experiments can be compromised is by misconduct and sloppy science within various stages of the empirical cycle (Simmons, Nelson & Simonsohn, 2011). The data collection and analysis phase as well as the reporting phase are most vulnerable to distortion by fraud, p-hacking and other questionable research practices (QRPs).

  • In the data collection phase, observations that (if kept) would lead to undesired conclusions or non-significant results can be altered or omitted. Also, fake observations can be added (fabricated).
  • In the analysis of data researchers can try alternative specifications of the variables, scale constructions, and regression models, searching for those that ‘work’ and choosing those that reach the desired conclusion.
  • In the reporting phase, things go wrong when the search for alternative specifications and the sensitivity of the results with respect to decisions in the data analysis phase is not disclosed.
  • In the peer review process, there can be pressure from editors and reviewers to cut reports of non-significant results, or to collect additional data supporting the hypotheses and the significant results reported in the literature.

Results from these forms of QRPs are that null-findings are less likely to be published, and that published research is biased towards positive findings, confirming the hypotheses, published findings are not reproducible, and when a replication attempt is made, published findings are found to be less significant, less often positive, and of a lower effect size (Open Science Collaboration, 2015).

Alarm bells, red flags and other warning signs

Some of the forms of misconduct mentioned above are very difficult to detect for reviewers and editors. When observations are fabricated or omitted from the analysis, only inside information, very sophisticated data detectives and stupidity of the authors can help us. Also many other forms of misconduct are difficult to prove. While smoking guns are rare, we can look for clues. I have developed a checklist of warning signs and good practices that editors and reviewers can use to screen submissions (see below). The checklist uses terminology that is not specific to experiments, but applies to all forms of data. While a high number of warning signs in itself does not prove anything, it should alert reviewers and editors. There is no norm for the number of flags. The table below only mentions the warning signs; the paper version of this blog post also shows a column with the positive poles. Those who would like to count good practices and reward authors for a higher number can count gold stars rather than red flags. The checklist was developed independently of the checklist that Wicherts et al. (2016) recently published.

Warning signs

  • The power of the analysis is too low.
  • The results are too good to be true.
  • All hypotheses are confirmed.
  • P-values are just below critical thresholds (e.g., p<.05)
  • A groundbreaking result is reported but not replicated in another sample.
  • The data and code are not made available upon request.
  • The data are not made available upon article submission.
  • The code is not made available upon article submission.
  • Materials (manipulations, survey questions) are described superficially.
  • Descriptive statistics are not reported.
  • The hypotheses are tested in analyses with covariates and results without covariates are not disclosed.
  • The research is not preregistered.
  • No details of an IRB procedure are given.
  • Participant recruitment procedures are not described.
  • Exact details of time and location of the data collection are not described.
  • A power analysis is lacking.
  • Unusual / non-validated measures are used without justification.
  • Different dependent variables are analyzed in different studies within the same article without justification.
  • Variables are (log)transformed or recoded in unusual categories without justification.
  • Numbers of observations mentioned at different places in the article are inconsistent. Loss or addition of observations is not justified.
  • A one-sided test is reported when a two-sided test would be appropriate.
  • Test-statistics (p-values, F-values) reported are incorrect.

With the increasing number of retractions of articles reporting on experimental research published in scholarly journals the awareness of the fallibility of peer review as a quality control mechanism has increased. Communities of researchers employing experimental designs have formulated solutions to these problems. In the review and publication stage, the following solutions have been proposed.

  • Access to data and code. An increasing number of science funders require grantees to provide open access to the data and the code that they have collected. Likewise, authors are required to provide access to data and code at a growing number of journals, such as Science, Nature, and the American Journal of Political Science. Platforms such as Dataverse, the Open Science Framework and Github facilitate sharing of data and code. Some journals do not require access to data and code, but provide Open Science badges for articles that do provide access.
  • Pledges, such as the ‘21 word solution’, a statement designed by Simmons, Nelson and Simonsohn (2012) that authors can include in their paper to ensure they have not fudged the data: “We report how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study.”
  • Full disclosure of methodological details of research submitted for publication, for instance through psychdisclosure.org is now required by major journals in psychology.
  • Apps such as Statcheck, p-curve, p-checker, and r-index can help editors and reviewers detect fishy business. They also have the potential to improve research hygiene when researchers start using these apps to check their own work before they submit it for review.

As these solutions become more commonly used we should see the quality of research go up. The number of red flags in research should decrease and the number of gold stars should increase. This requires not only that reviewers and editors use the checklist, but most importantly, that also researchers themselves use it.

The solutions above should be supplemented by better research practices before researchers submit their papers for review. In particular, two measures are worth mentioning:

  • Preregistration of research, for instance on Aspredicted.org. An increasing number of journals in psychology require research to be preregistered. Some journals guarantee publication of research regardless of its results after a round of peer review of the research design.
  • Increasing the statistical power of research is one of the most promising strategies to increase the quality of experimental research (Bakker, Van Dijk & Wicherts, 2012). In many fields and for many decades, published research has been underpowered, using samples of participants that are not large enough the reported effect sizes. Using larger samples reduces the likelihood of both false positives as well as false negatives.

A variety of institutional designs have been proposed to encourage the use of the solutions mentioned above, including reducing the incentives in careers of researchers and hiring and promotion decisions for using questionable research practices, rewarding researchers for good conduct through badges, the adoption of voluntary codes of conduct, and socialization of students and senior staff through teaching and workshops. Research funders, journals, editors, authors, reviewers, universities, senior researchers and students all have a responsibility in these developments.

References

Bakker, M., Van Dijk, A. & Wicherts, J. (2012). The Rules of the Game Called Psychological Science. Perspectives on Psychological Science, 7(6): 543–554.

Henrich, J., Heine, S.J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33: 61 – 135.

Ioannidis, J.P.A. (2005). Why Most Published Research Findings Are False. PLoS Medicine, 2(8): e124. http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.0020124

Kerr, N.L. (1989). HARKing: Hypothesizing After Results are Known. Personality and Social Psychology Review, 2: 196-217.

Open Science Collaboration (2015). Estimating the Reproducibility of Psychological Science. Science, 349. http://www.sciencemag.org/content/349/6251/aac4716.full.html

Shadish, W.R., Cook, T.D., & Campbell, D.T. (2002). Experimental and quasi-experimental designs for generalized causal inference. Boston, MA: Houghton Mifflin.

Simmons, J.P., Nelson, L.D., & Simonsohn, U. (2011). False positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22: 1359–1366.

Simmons, J.P., Nelson, L.D. & Simonsohn, U. (2012). A 21 Word Solution. Available at SSRN: http://ssrn.com/abstract=2160588

Wicherts, J.M., Veldkamp, C.L., Augusteijn, H.E., Bakker, M., Van Aert, R.C & Van Assen, M.L.A.M. (2016). Researcher degrees of freedom in planning, running, analyzing, and reporting psychological studies: A checklist to avoid p-hacking. Frontiers of Psychology, 7: 1832. http://journal.frontiersin.org/article/10.3389/fpsyg.2016.01832/abstract

1 Comment

Filed under academic misconduct, experiments, fraud, incentives, open science, psychology, survey research

Four Reasons Why We Are Converting to Open Science

The Center for Philanthropic Studies I am leading at VU Amsterdam is converting to Open Science.

Open Science offers four advantages to the scientific community, nonprofit organizations, and the public at large:

  1. Access: we make our work more easily accessible for everyone. Our research serves public goods, which are served best by open access.
  2. Efficiency: we make it easier for others to build on our work, which saves time.
  3. Quality: we enable others to check our work, find flaws and improve it.
  4. Innovation: ultimately, open science facilitates the production of knowledge.

What does the change mean in practice?

First, the source of funding for contract research we conduct will always be disclosed.

Second, data collection – interviews, surveys, experiments – will follow a prespecified protocol. This includes the number of observations forseen, the questions to be asked, measures to be included, hypotheses to be tested, and analyses to be conducted. New studies will be preferably be preregistered.

Third, data collected and the code used to conduct the analyses will be made public, through the Open Science Framework for instance. Obviously, personal or sensitive data will not be made public.

Fourth, results of research will preferably be published in open access mode. This does not mean that we will publish only in Open Access journals. Research reports and papers for academic will be made available online in working paper archives, as a ‘preprint’ version, or in other ways.

 

December 16, 2015 update:

A fifth reason, following directly from #1 and #2, is that open science reduces the costs of science for society.

See this previous post for links to our Giving in the Netherlands Panel Survey data and questionnaires.

 

July 8, 2017 update:

A public use file of the Giving in the Netherlands Panel Survey and the user manual are posted at the Open Science Framework.

1 Comment

Filed under academic misconduct, Center for Philanthropic Studies, contract research, data, fraud, incentives, methodology, open science, regulation, survey research

Een gedragscode voor werving nalatenschappen door goede doelen

In 1Vandaag zei ik op 19 februari dat er in Nederland geen gedragscode voor de werving van nalatenschappen bestaat. Dit blijkt niet waar, er is wel degelijk een richtlijn voor nalatenschappenwerving. In de regels van het Centraal Bureau Fondsenwerving voor het CBF-Keur en op de website van de VFI (Vereniging voor Fondsenwervende Instellingen), branchevereniging voor goede doelen, is deze richtlijn echter niet te vinden. De VFI heeft wel een richtlijn voor de afwikkeling van nalatenschappen. Maar die gaat over de afwikkeling, als het geld al binnen is. Niet over de werving van nalatenschappen.

Het blijkt dat de richtlijn voor de werving van nalatenschappen is gepubliceerd door een derde organisatie, het Instituut Fondsenwerving. Deze organisatie heeft in 2012 een richtlijn opgesteld voor fondsenwervende instellingen die nalatenschappen werven. De richtlijn is niet verplichtend. Het Instituut Fondsenwerving heeft ook een gedragscode waar haar leden zich aan hebben te houden, maar de richtlijn voor nalatenschappen heeft niet de status van gedragscode. Bij het Instituut Fondsenwerving zijn volgens de ledenlijst 231 goede doelen organisaties aangesloten (klik hier voor een overzicht in Excel). Het Leger des Heils is lid van het IF, maar de Zonnebloem niet. Ook andere grote ontvangers van nalatenschappen, zoals KWF Kankerbestrijding, ontbreken op de ledenlijst. Zij zijn wel lid van de VFI, dat 113 leden en 11 aspirant leden telt.

De VFI reageerde op de uitzending via haar website en vermeldde de richtlijn van het Instituut Fondsenwerving. De status van de richtlijn is in de reactie opgehoogd naar een gedragscode. Dit zou betekenen dat leden die zich niet aan de richtlijn houden, kunnen worden geroyeerd. Gosse Bosma, directeur van de VFI, zei overigens in de 1Vandaag uitzending dat de VFI niet is nagegaan of de betrokken leden zich aan de richtlijn hebben gehouden en dat ook niet nodig te vinden. Het IF zelf heeft niet gereageerd. Ook de ontvangende goede doelen, de Zonnebloem en het Leger des Heils, reageerden deze week via het vaktijdschrift voor de filantropie, Filanthropium. Zij verklaarden zich bereid onrechtmatigheden te corrigeren. Wordt ongetwijfeld vervolgd in de volgende fase van deze zaak, of wanneer een nieuwe zaak zich aandient.

 

Leave a comment

Filed under bequests, charitable organizations, fundraising, incentives, law, philanthropy, regulation

Wat zegt het CBF-Keur voor goede doelen?

Het Financieel Dagblad besteedt een lang artikel aan de betekenis van het CBF-Keur voor goede doelen naar aanleiding van de vraag: “Waar blijft mijn gedoneerde euro?” Het “keurmerk en boekhoudregels zijn geen garantie voor een zinvolle besteding”, volgens de krant. Verderop in het artikel staat mijn naam genoemd bij de stelling dat het CBF-Keur ‘fraude of veel te hoge kosten niet uitsluit’ en zelfs dat het ‘nietszeggend’ zou zijn. Inderdaad zegt het feit dat een goed doel over het CBF-Keur beschikt niet dat de organisatie perfect werkt. Het maakt fraude niet onmogelijk en dwingt organisaties ook niet altijd tot de meest efficiënte besteding van beschikbare middelen. In het verleden zijn misstanden bij verschillende CBF-Keurmerkhouders in het nieuws gekomen, die bij sommige organisaties hebben geleid tot intrekking van het keurmerk.

Maar helemaal ‘nietszeggend’ is het CBF-Keur ook weer niet. Zo denk ik er ook niet over. Het CBF-Keur zegt wel degelijk wat. Voordat een organisatie het CBF-Keur mag voeren moet het een uitgebreide procedure door om aan eisen te voldoen aan financiële verslaggeving, onafhankelijkheid van het bestuur, kosten van fondsenwerving, en de formulering van beleidsplannen. Dit zijn relevante criteria. Zij zorgen ervoor dat je als donateur erop kunt vertrouwen dat de organisatie op een professionele manier werkt. Het CBF-Keur zegt alleen niet zoveel over de efficiëntie van de bestedingen van een goed doel. Veel mensen denken dat wel, zo constateerden we in onderzoek uit 2009.

Het is lastige materie. Garantie krijg je op een product dat je koopt in de winkel, waardoor je het terug kunt brengen als het niet functioneert of binnen korte tijd stuk gaat. Zulke garanties zijn moeilijk te geven voor giften aan goede doelen. Een dergelijke garantie zou je alleen kunnen geven als de kwaliteit van het werk van goede doelenorganisaties gecontroleerd kan worden en er een minimumeis voor te formuleren valt. Dat lijkt mij onmogelijk. Het CBF-Keur is niet zoiets als een rijbewijs dat je moet hebben voordat je een auto mag besturen. De markt voor goede doelen is vrij toegankelijk; iedereen mag de weg op. Sommige goede doelen hebben een keurmerk, maar dat zegt vooral hoeveel ze betaald hebben voor de benzine, in wat voor auto ze rijden en wie er achter het stuur zit. Het zegt nog niet zoveel over de hoeveelheid ongelukken die ze ooit hebben mee gemaakt of veroorzaakt, en of dat de kortste of de snelste weg is.

Vorig jaar stelde de commissie-De Jong voor om een autoriteit filantropie in te stellen, die organisaties zou gaan controleren voordat ze de markt voor goede doelen op mogen. Er zou een goede doelen politie komen die ook op de naleving van de regels mag controleren en boetes mag uitdelen. Dat voorstel was te duur voor de overheid. Voor de goede doelen was het onaantrekkelijk omdat zij aan nieuwe regels zouden moeten gaan voldoen. Bovendien was het niet duidelijk of die nieuwe regels ook echt het aantal ongelukken zou verlagen. Het is op dit moment überhaupt niet duidelijk hoe goed de bestuurders van goede doelen de weg kennen en hoeveel ongelukken ze maken. Een beter systeem zou moeten beginnen met een meting van het aantal overtredingen in het goede doelen verkeer en een telling van het aantal bestuurders met en zonder rijbewijs. Vervolgens zou het goed zijn om een rijopleiding op te zetten die iedereen die de markt op wil kan volgen en in staat stelt de vaardigheden op te doen waarover elke bestuurder moet beschikken. Ik hoop dat het artikel in het Financieel Dagblad tot een discussie leidt die dit duidelijk maakt.

Intussen heeft het CBF gereageerd met de verzekering dat er gewerkt wordt aan uitwerking van richtlijnen voor ‘reactief toezicht op prestaties’. Ook de VFI, branchevereniging voor goede doelen, kwam met een reactie van die strekking. Dat is goed nieuws. Maar die nieuwe richtlijnen zijn er nog lang niet. In de tussentijd geeft het CBF keurmerken af en publiceren de Nederlandse media – die na Finland de meest vrije ter wereld zijn – af en toe een flitspaalfoto van wegmisbruikers. Dat lijkt voldoende te zijn om het goede doelen verkeer zichzelf te laten regelen en de ergste ongelukken te voorkomen. Want die zijn er maar weinig.

Leave a comment

Filed under charitable organizations, fraud, household giving, impact, incentives, law, philanthropy, policy evaluation, politics

Conditional Review Acceptance Policy

Here’s CRAP: a new policy regarding review requests I’ve decided to try out. CRAP means Conditional Review Acceptance Policy, the new default response to review requests. I will perform review only if the journal agrees to publish the article in a Free Open Access mode – making the article publicly available, without charging any fees for it from universities, authors, or readers.

Here’s the story behind CRAP. If you’re an academic, you will recognize the pattern: you get an ‘invitation’ or a ‘request’ to review a paper submitted to the journal because ‘you have been identified as an expert on the topic’. If you’re serious about the job, you easily spend half a day reading the article, thinking about the innovations in the research questions, the consistency of the hypotheses, wondering why previous research was ignored, vetting the reliability and the validity of the data and methods used, checking the tables, leaving aside the errors in references which the author copied from a previously published article. As a reviewer accepting the task to review a paper you sometimes get a 25% discount on the hugely overpriced books by the publisher or access to journal articles which your university library already paid for.

You accept the invitation because you know the editor personally, you want to help improve science, want to facilitate progress in the field, because by refusing you will miss an opportunity to influence the direction your field is taking or simply to block rubbish from being published. *Or, if you have less laudable objectives, because you want others to know and cite your work. I confess I have fallen prey to this temptation myself.* But it does not end after the job is done. There’s a good chance you will get the article back after the authors revised it, and you are invited again to check whether the authors have done a good job incorporating your comments. In the mean time, you’ve received seven more review requests. I could fill my entire week reviewing papers if I accepted all invitations I receive.

In the world outside academia complying with a request means doing people a favor, which at some point in the future you can count on to be returned. Not so in academia. The favors that we academics are doing are used by publishers to make profit, by selling the journals we work for as unpaid volunteers to university libraries. The journal prices that publishers are ridiculously high but libraries have no choice but to accept them because they cannnot afford to miss the journal in its collection. And ultimately we keep up the system by continuing to accept review requests. Academic publishers exploit scholars asking for reviews and giving nothing in return.

If you’re not an academic you may find this all very strange. When I told my parents in 2003 that my first article was accepted for an international journal they asked: “How much did they pay you for the article?” Journalists and free lance writers for magazines may get paid for the content they are producing, but not academics. The content we hope to help produce as reviewers is a public good: valid and reliable knowledge. Falsity should be avoided at all cost; the truth, the truth, and nothing but the truth should be published. The production of this public good is facilitated by public money. But the reviews we provide are not public goods. They are private goods. They are typically anonymous, and not shared publicly. We send them to the editorial assistant, who sends them to the authors (and sometimes, ‘as a courtesy’, to the other volunteer reviewers). The final product is again a private good, sold by the publisher. Collectively, our favors are creating a public bad: increasing costs for journal subscriptions.

What can we do about this? Should the volunteer work we do be monetized? Should we go on strike to ask for an adequate wage? According to the profitability of the journal perhaps? So that the higher the profit the publisher makes on a journal, the higher the compensation for reviewers? This would do nothing to reduce the public bad. Instead, I think we should move into Free Open Access publishing. The public nature of knowledge, the production of which is made possible by public funding, should be accessible for the public. It is fair that some compensation is given to the journal’s publisher for the costs they will have to incur to copy-edit the article and to host the electronic manuscript submission system. These costs are relatively low. I am leaving the number crunching for some other time or some other geek, but my hunch is that if we would monetize our volunteer work as reviewers this would be enough to pay for the publication of one article.

Academic publishers are not stupid. They see the push towards open access coming, and are now actively offering open access publication in their journals. But everything comes at a price. So they are charging authors (i.e., authors’ funders) fees for open access publication, ranging from several hundred to thousands of dollars. Obviously, this business model is quite profitable – otherwise commercial publishers would not adopt it. Thugs and thieves are abusing the fee-based open access model by creating worthless journals that will publish any article, cashing the fees and to make a profit. The more respectable publishers are now negotiating with universities and public funders of science about a better model, circumventing the authors. Undoubtedly the starting point for such a model is that the academic publication industry remains profitable. In all of this, the volunteer work of reviewers is still the backbone of high quality journal publications. And it is still not compensated.

So my plea to fellow academics is simple. We should give CRAP as our new default response. Agree to review if the publisher agrees to publish the article in Free Open Access. It may be the only way to force Free Open Access into existence. I will keep you updated on the score.

Update: 14 October 2014
Invited: 8
Response: 4 Declined (Journal of Personality and Social Psychology, Sociology of Religion, Nonprofit and Voluntary Sector Quarterly, Science and Public Policy; Qualitative Sociology); 1 offered Green Access (Public Management Review); 2 responded review was no longer needed (Journal for the Scientific Study of Religion and Body & Society).

One managing editor wrote: “Thank you for offering to review this manuscript. Unfortunately, our publisher has not yet approved free Open Source. Those of us who actually work for the journal instead of the publishing institution would gladly provide open access to articles if it was up to us. These kinds of decisions are not left up to our discretion however. I greatly respect your stance and hope it is one that will eventually lead to greater access to academic publications in the future.”

In a message titled “Your assignment”, the associate editor of JPSP wrote:  “I appreciate your willingness to review manuscript #[omitted] for Journal of Personality and Social Psychology:  Personality Processes and Individual Differences. As it turns out, your review will not be needed for me to make a decision, so please do not complete your review. ”

The message from the editor of Sociology of Religion, probably composed by an Oxford University Press employee, says: “Sociology of Religion does not have an author-pays Open Access option in place, which would require the author or the body that funded the research to pay an Author Processing Charge—there is a range of APCs, beginning at $1,800. This is the only system currently in place at Oxford University Press for optional Open Access (some journals, of course, are entirely Open Access by design, generally with significant society sponsorship). The request and APC would need to come from the author, not the manuscript reviewer. Moreover, if an author requires OA to comply with requirements from his or her funding body, then the author submits it to a journal that has a OA option. Also, all authors of published articles are given a toll-free URL to post wherever they like—this allows the final version to be read without payment by anyone using that link, and importantly, counts toward online usage statistics. While this isn’t exactly the same as OA, it does make it freely available through that link as it is posted or distributed by the author.”

This is an interesting response from OUP. The question aside why the ‘Author Processing Charge’ must be as high as $1,800, if three reviewers each charge $600 for the volunteer work they provide for the journal by reviewing the paper, the APC would be compensated. As a courtesy to reviewers, OUP could waive the APC. Reviewers could wave the review fee as a courtesy to OUP. With wallets closed everybody benefits.

The editor of Science and Public Policy, another OUP journal, responded: “unfortunately at present an unconditional policy for open access publishing is not in place for our journal, rather the following policy applies, which is not in line with your conditions, according to which Authors may upload their accepted manuscript PDF to an institutional and/or centrally organized repository, provided that public availability is delayed until 24 months after first online publication in the journal.”

The editor of the Journal for the Scientific Study of Religion wrote in what seems to be a standard reply: “It has become apparent that I will not need you to review the manuscript at this time. I hope you will be able to review other manuscripts for JSSR in the near future.”

In contrast, the editorial assistant for Body & Society wrote: “We’re just writing to let you know that we no longer require you to review this paper. Enough reviews have come in for the editorial board to be able to make a decision. Thank you for having agreed to review, and we apologise for any inconvenience caused.”

*HT to @dwcorne for identifying this less benign motivation.

11 Comments

Filed under incentives, volunteering