Voices against attempts to crush freedom of research at the University of Cape Town

The University of Cape Town is a fine university – one I was proud to have attended. It is the highest rated university in Africa, even though it has dropped from number 191 in the ranking in 2018 to number 220 today.

That reputation has been hard won. It took years of resisting the apartheid government, although not always successfully. The independence of its academics to pursue their research was always essential to this endeavour.

Now there is an attempt to crush that independence and with it the academic freedom so essential to university life.

Below are more voices raised in protest against what has taken place. You can find the origins of this campaign here, with further contributions here, here, here, here, here, here, here, here, here and here.

It is because the issues are so critical that they have resulted in such passionate defences of our university’s precious heritage.

Martin

---

Open Letter to Deputy Vice-Chancellor Harrison on her role in the ‘witch-hunt’ against Professor Nicoli Nattrass, 19th June 2020

Dear Deputy Vice-Chancellor Harrison,

I believe that your actions over the past three weeks reveal multiple conflicts of interest and inconsistency on your part.

I have sent you a series of emails and detailed private letters posing questions concerning your leading role in what I am experiencing as an unprocedural and prejudicial ‘witch-hunt’ against me. In raising issues with you in private I showed you the courtesy that you failed to show me before you rushed to judgement in condemning publicly me and my research. You have acknowledged receipt of my letters but chosen not to answer my questions about the current process of investigation into me because, you say, the process is underway. I have tried other means of addressing what I am experiencing as an abuse of power but UCT’s procedures appear to have broken down. Given your refusal to engage with me in private and the inadequacy of formal channels at UCT, I am now writing an Open Letter summarizing my concerns over your role.

On Friday 5th June, UCT issued a public statement from its ‘Executive’ passing summary judgement on me and my Commentary (published in the SA Journal of Science). The statement named me twice. The University tweeted this to its 207,000 followers. Last Friday (12th June), the Vice-Chancellor (VC) told Senate that the ‘Executive’ comprised herself, the three Deputy VCs and two executive officers. You are DVC for Research and Internationalisation (DVC-R). Whilst your responsibilities do not appear to be set out on the UCT website, it is reasonable to assume that you take primary responsibility for the oversight of research at the university. Given that the statement from the UCT ‘Executive’ concerned research, you presumably played a leading role in crafting it. Indeed, the VC told UCT’s Senate on Friday that ‘When I was inundated with emails and whatever, I actually requested Professor Harrison to lead, to look at the paper … I asked her to lead in the conversation on this because it is in her portfolio.’ The VC appears to be saying that you played the leading role in the discussion that led to the statement from the ‘Executive’. One might describe you as its architect. The VC also told Senate that there had been no prior investigation other than what she had asked you to do as DVC-R.

You yourself are a highly-regarded scholar. I surely know less about your field (biomass engineering) than you know about mine (social science), but I can see that your scholarship is highly cited (you have an h-index of 37) and you have won numerous impressive accolades. You were the founding director of UCT’s highly-regarded Future Water research institute which is a very impressive initiative. Less than a year ago you were appointed as DVC-R. I myself welcomed your appointment, delighted that a serious scholar and esteemed colleague would be at the helm of research at the University.

In the Executive statement you judged that my research suffered ‘conceptual and methodological flaws’ and implied that it was ethically deficient. I have repeatedly asked you what are these flaws and you have repeatedly declined to elaborate. 2

The Executive statement of 5th June also condemned me and my Commentary for ‘unexamined assumptions about what black people think …’. The statement continued: ‘The paper is offensive to … black people in general’ (among others). As other scholars have already noted, your statement condemns me (incorrectly, in my view) for implicit generalization when you yourself engage in explicit generalization. As the clear expert on research within the ‘Executive’, what evidence you relied on when making this generalization?

As DVC-R, you appear to have rushed to judgement on me and my Commentary and – as the expert on research within the Executive – been the architect of an apparently unprecedented and totally inappropriate public statement of censure in advance of any substantive investigation. Your strong and very public condemnation surely further ‘poisoned the well’ for any subsequent investigation. All allegations against me for Research Misconduct have been formulated and assessed in the light of the outcry from some people which the UCT Executive fueled with its public statement.

What makes this especially troubling is that you yourself had extensive prior knowledge of my mini-study and my findings. You will recall that last December you yourself chaired an institutional review panel where my apparently flawed questions and findings were presented. You also received documents in advance where the findings were summarized. You will recall that I presented the principal findings in a series of Powerpoint slides. These showed clearly the variables used, descriptive statistics and the results of the exploratory multi-variate regression models. These were pretty much what was published in the SAJS Commentary. You will recall that no one at the panel meeting raised any concerns over the questions or the analysis. You yourself did not raise any concerns. You did not suggest that my mini-study had conceptual and methodological flaws or that it was ethically deficient. On the contrary, the external members of the review panel were then and later, in their report sent to you, complementary about the exploratory research. One of the external reviewers recommended – in your presence – that the findings be published.

I am astonished that you sat through the panel meeting, read the relevant documents circulated in advance and the reviewers’ report circulated subsequently, and – over a period of six months – chose not to raise any concern over what you now apparently believe to be obvious flaws in my research. At the very least, you appear to be inconsistent.

A cynic might wonder whether you have an interest in trying to ‘bury’ this record or your prior knowledge of and evident acquiescence in such allegedly flawed research.

I would now like to turn to your role in what I am experiencing as a ‘witch-hunt’ against me subsequent to your public condemnation of me and my research in the statement of 5th June.

You yourself are presiding over the Research Misconduct investigation into me. UCT has a well-established set of procedures for dealing with such complaints, designed to protect both the people making the complaints and the researchers against whom the complaints were made. Research ethics processes happen in the first instance at the faculty level, in part to ensure that they take into account subject knowledge and ethical codes that are often specific to an academic 3

discipline. Almost all social science research, for example, is considered within the Commerce and Humanities faculties, where scholars understand the social sciences and the research ethics codes and principles that are most relevant to them.

My understanding is that, when complaints began to arrive on Thursday 4th June, UCT’s Office for Research Integrity correctly forwarded them to the Commerce Faculty, i.e. the faculty where I have an academic appointment and which approved my research ethics application in 2019. What happened then is not entirely clear to me. It appears that, in discussion with the Commerce Faculty, you yourself wrestled back control over the investigation. I understand that this was on Saturday 6th June, more than 24 hours after the Executive had issued its public statement of condemnation. I believe that this was unprocedural in terms of Commerce Faculty and University policies.

This is extraordinary. How can the DVC-R who was the architect of the prior public condemnation of me and my Commentary then proceed to preside over a subsequent investigation? I have repeatedly asked you this in private correspondence. You have repeatedly declined to answer.

I have been told that you yourself then ‘initiated’ the first, ‘informal’ stage of a Research Misconduct investigation. You yourself approached two assessors without any obvious expertise in social survey research. You then directed them to investigate two ‘charges’ which, in my view, indicate a lack of understanding of the ethical approval processes governing social survey research and of the relevant ethical codes and practices in my Faculty. This is not the place to discuss these charges (which I believe are baseless). My concern here and now is your role in this. You, the architect of a public statement of condemnation then proceeded to direct the investigation down a particular path, in violations of university procedures, in what appears to me to be an attempt to find some post-hoc justification for your ill-considered Statement. I have repeatedly asked you about this in private correspondence. You have repeatedly declined to answer.

I have grave doubts that any investigation at UCT can be fair given your statement of 5th June and the controversy that it fuelled. Your statement, to put it bluntly, ‘poisoned the well’. The well has, of course, been further poisoned by statements and newspaper articles from various other university officers.

As you will know from my published articles on News24, I am deeply concerned over the threat that actions of the Executive have posed to the university. I am not going to repeat what I wrote there, except to reiterate that I am worried about the university being subjected to thought-police. I now worry about your role in this. It seems to me that you are presiding over a dangerous shift in how research is policed at UCT. Your investigation is shifting the focus of research ethics review from research – i.e. the collection of data – to the analysis and interpretation of that data in a publication. It seems to me that you are taking the university down a path that will lead to the requirement that all researchers submit all articles and books to a research ethics office prior to publication, lest the analysis and interpretation causes offence to someone. You will be aware that some of the attacks on me in the media have explicitly called for the university to establish an office for the ‘vetting’ of all papers prior to publication. 4

University policies (including Research Misconduct policies) emphasise again and again that ‘conflicts of interest must be avoided’. You seem to have proceeded without any concern over your conflicts of interest. Having been present when the detailed research findings were presented and publication recommended, and having read documents summarising and praising the research, you took the lead in the discussion within the Executive leading to the public statement of 5th June. This is a statement that the VC herself has now described as incorrect in some respects and which the Executive is ‘reconsidering’. Having taken the leading role in the flawed process leading to a flawed statement, you wrestled control over the Research Misconduct investigation away from the responsible faculty, in violation of the university’s procedures. You then initiated an investigation into me, with what can reasonably be seen as the objective of providing post hoc justification for your prior statement. You appointed investigators with no evident experience of social science survey research and it seems that you yourself directed them to focus on specific charges. At every stage the process appears compromised.

Is this conduct appropriate for a DVC for Research at a prestigious university?

I cannot understand your actions. I cannot understand why you did not step aside three weeks ago. I cannot understand why you have not now resigned.

Yours

Professor Nicoli Nattrass.

---

Letter to the DVC-R and VC re the charge of (deliberate) falsification against Professor Nattrass

 

I am writing to you about the charges levelled against my colleague, Professor Nicoli Nattrass (of UCT’s School of Economics and the Centre for Social Science Research, CSSR). You will know well that she is also my wife and I am assisting her as best I can. I am not writing as her husband, however. I am writing in my capacity as a senior social scientist, as a founding member of the CSSR and – for the past ten years – the CSSR’s director.

 

I am aware that the Deputy Vice-Chancellor for Research (DVC-R) appointed a Preliminary Informal Enquiry into complaints of ‘Research Misconduct’ by Nicoli, arising from the publication of a 2-page Commentary in the South African Journal of Science. I appreciate that some of the criticisms of the Commentary raise difficult and important issues that require appropriate reflection. I acknowledge that the Vice-Chancellor herself has conceded that the university mishandled its response. I know also that Nicoli has asked a number of questions about the Research Misconduct process being followed, including why this investigation was ‘escalated’ to the DVC-R rather than being conducted through the Commerce Faculty and according to Commerce Faculty policies and procedures. Leaving aside all of these issues, I wish to focus on one specific issue: The charge levelled against Nicoli of ‘falsification’. I believe that this charge is entirely inappropriate and poses a grave threat to social science survey research at the university and to the university’s global reputation.

 

I understand that Nicoli is being investigated for ‘falsification’ in relation to (a) her application for research ethics approval through the Commerce Faculty to conduct a mini-survey of students (an application that was approved in 2019) and (b) the consent forms signed by the students who were the respondents or interviewees in her mini-survey. The essence of the charge is that Nicoli did not specify that she would write a Commentary on student subject choice (and whether this helped to explain the small numbers of black South African students in biological sciences and conservation biology in particular).

 

This is an amazing charge. To the best of my knowledge, no social science survey research application or consent form specifies the papers that will be written using the data. Research ethics processes with respect to social science survey research focus on the collection of data (and their responsible storage). Research ethics processes have never been concerned with the papers (or commentaries) written in future years using the data.

 

If upheld, this charge against Nicoli would shut down most social science survey research at UCT and would prevent survey data being put in the public domain (as required by many funders and now UCT itself) – because there would be no control over the uses to which the data could be put, even if the data collection (and storage) had been approved.

 

The charge of falsification is set out in paragraph 3.2 of the UCT Policy and Procedures for Breach of Research Ethics Codes and Allegations of Misconduct in Research (i.e. Research Misconduct Policy):

3.2 Falsification – deliberate misrepresentation of research including progress in

research or inappropriate adjustment and/or selection of data, imagery, results and/or

consents, or undisclosed duplication of publication, or inappropriate claims to authorship or

attribution of work contrary to the UCT Authorship Practices Policy

This identical paragraph 3.2 is included in the ‘Commerce Faculty Policy and Procedures for Breach of Research Ethics Codes and Allegations of Misconduct in Research’.

 

You may be aware that I was integrally involved in writing UCT’s research ethics policies that were approved by Senate in the early 2010s. I have no recollection, however, of the discussion over this. I imagine that it was copied from global codes of research ethics. I myself have had no previous experience of cases of alleged ‘falsification’.

A simple google search for ‘falsification research ethics’ points to a publication from the US National Academy of Science, titled On Being a Scientist: A Guide to Responsible Conduct in Research (3^rd edition, https://www.ncbi.nlm.nih.gov/books/NBK214564/). We have not accessed the entire book, but parts are extracted on the internet. Falsification is described or defined as ‘manipulating research materials, equipment, or processes, or changing or omitting data or results such that the research is not accurately represented in the research record’. The online extract goes on to say that ‘a crucial distinction between falsification, fabrication, and plagiarism (sometimes called FFP) and error or negligence is the intent to deceive’.

The emphasis on deliberate misrepresentation of research is included in the UCT policy documents.

I would like to suggest that no serious scholar in the quantitative social sciences would believe that Nicoli deliberately omitted from her research ethics application and consent form any mention of an intention to write about student subject choice. The charge of deliberate deception smacks of a malicious prosecution.

This charge against Nicoli has very worrying implications for social science survey research at UCT.

The university’s research ethics processes are based on the devolution of responsibility to the faculties. Research ethics processes happen in the first instance at the faculty level, in part to disperse the administrative burden and in part to ensure that they take into account subject knowledge and ethical codes that are often specific to an academic discipline. Almost all social science survey research, for example, is considered within the Commerce Faculty (and the Centre for Social Science Research, now in the Humanities Faculty), where scholars understand the social sciences, their methodologies and the research ethics codes and principles that are most relevant to them.

The charge of falsification ignores entirely the norms and standards of research ethics for social science survey research.

The US National Academy of Sciences website (see above) has this to say about a US ‘federal statement’:

‘… the federal statement says that to be considered research misconduct, actions must represent a “significant departure from accepted practices,” must have been “committed intentionally, or knowingly, or recklessly,” and must be “proven by a preponderance of evidence.” According to the statement, “research misconduct does not include differences of opinion.”’

What are ‘accepted practices’ in social science survey research at UCT and elsewhere?

Research ethics applications are, in general, confidential. In my experience – on Senate, faculty and departmental research ethics committees – applications for research ethics ‘clearance’ for social science survey research have never specified or been asked to specify the papers that would be written using the data in years to come.

Nor have the consent forms signed by respondents specified the papers that would be written using the data in years to come. To do so might even constitute ‘priming’ respondents by framing their understanding of the questions asked in the survey.

Consider the following examples of consent forms that were approved through the Commerce Faculty or CSSR at UCT:

‘This is a study about household welfare in South Africa, run by researchers at the University of Cape Town for the South African Presidency. The purpose of this study is to learn more about how people in South Africa are faring over time … We would like to ask you some questions about your education, employment, income, family and health.’ (NIDS, wave 5, adult questionnaire)

‘South Africa is a young democracy and we need to understand your experience of our May 2019 election.  We are doing this study for the Afrobarometer, a pan-African network of researchers who study issues of governance, democracy, elections, markets in over thirty African countries, to learn more about citizens’ views on politics, economics and related issues. … This information will help us understand how South African citizens view the current political and economic situation and what needs to be done to make them better.  Our analysis of this information will be made available to members of the public through the media, civil society, government officials and international development partners. The data will be used for research by university scholars.’ (Afrobarometer, round 8, 2019)

‘We are conducting a national survey of South Africans for research purposes for the Institute for Justice and Reconciliation, a non-profit organisation in Cape Town, South Africa. …’ (Reconciliation Barometer, 2020)

These (and other) consent forms provide (at most) a broad range of topics that might be examined using the data collected through the surveys. The lists of topics are in no way exhaustive or exclusive.

What about social science surveys beyond UCT? Consider the following examples of surveys of attitudes and beliefs on very sensitive topics:

‘Hello, … I am from the Human Sciences Research Council and our organisation is asking people in South Africa and in your community to answer questions, which we hope will benefit South Africans in the future. We are conducting a national survey which aims to determine the public’s attitudes, beliefs, behaviour patterns and values with regard to various important issues affecting South Africa. We run this survey every year to see how these attitudes are changing over time. The questionnaire covers a wide range of topics, with examples including attitudes about democracy and governance, service delivery, race relations, crime, moral issues, and poverty.’ (South African Social Attitudes Survey, 2017)

‘We are looking for people aged 15 or over who are willing to share their opinions about current events and political or social issues for a study we are conducting across Europe.’ (Eurobarometer, 2019)

These datasets have been used for papers on very sensitive topics.

Global best practice and now UCT policy require that we archive the data used in publications. At UCT we have what is probably the premier social science data portal in the global South: DataFirst. Data from studies such as NIDS, Afrobarometer and the Reconciliation Barometer are archived and made publicly available through DataFirst (and other portals). Once data are thus archived, no one has any control over how the data will be used or for what purposes. All of these datasets include race variables. Scholars could use the data for controversial, even offensive purposes. The only way of preventing this would be to terminate public access or to police the use of the data by requiring that users of the data submit for vetting all papers prior to publication. Given that such policing would be completely unacceptable to the global community of social science scholars, we would have no choice but to shut down DataFirst and to discontinue most of our survey data-collection.

To say that this would have a chilling effect on social science research at UCT is of course a massive understatement. Moreover, it would expose UCT to global ridicule. The charge is entirely inappropriate for social science survey research and should never have been entertained in a Research Misconduct investigation.

 

I urge the university to commit itself publicly to the defence of research ethics on the basis of appropriate norms and standards.

 

Yours

Jeremy Seekings

Professor of Political Studies and Sociology,

Director, Centre for Social Science Research,

University of Cape Town.

---

In defence of Nicoli Nattrass: SA needs fearless debate, not a knee-jerk response to the Twitter-sphere