Hans Eysenck: controversialist or worse?

I heartily endorse the efforts of Smith, and Pelosi and Marks, as they shine the spotlight on the highly questionable collaborative work of Hans Eysenck and Ronald Grossarth-Maticek.[1] Playing with Fire: The Controversial Career of Hans J. Eysenck detailed the background to this unlikely alliance between Britain’s most famous psychologist and an isolated, outsider figure that few found credible.[2]

Smith recounts how Eysenck labelled the too-good-to-be-true criticism of his work with Grossarth-Maticek as unscientific and therefore unanswerable. This was indeed a clever deflection. As Pelosi noted, some of Grossarth-Maticek’s publications that predated Eysenck’s involvement featured results that were statistically impossible - and this was a scientific statement. Eysenck’s influence would turn the impossible into the merely unbelievable, with their joint publications featuring effect sizes and risk ratios an order of magnitude above any previously reported.

Eysenck did not simply act as a secondary guarantor for Grossarth-Maticek’s research program, as some still suggest. He played a key role in bringing Grossarth-Maticek’s work to the attention of English-speaking audiences. While Eysenck wrote several accounts of this research as a sole author, he seldom appeared as first author in co-authored publications. But I am certain he took the lead in writing them as well – as was his habit in his other collaborations. When I interviewed Grossarth-Maticek in 2003, we relied heavily on an interpreter. It seems implausible he had the capacity to write English to publication standards. Language barriers and the ‘international’ nature of this work also provided something of a shield. When pressed for answers in a 1991 target article symposium in Psychological Inquiry, Eysenck ultimately directed questions to Grossarth-Maticek.[3] This appeared to be another clever deflection, Eysenck knowing that it would be difficult to get clear answers from Grossarth-Maticek.

Eysenck’s collaboration with Grossarth-Maticek was very much a retirement project. Greater freedom allowed Eysenck to travel to Germany to frequently liaise with Grossarth-Maticek. Eysenck helped Grossarth-Maticek gain tobacco industry funding. With Eysenck’s input, idiosyncratic psychodynamic theorising was recast within a more mainstream typological framework. Clearly, this was collaborative research, which is why responsibility for an investigation rightfully rests with British authorities and should hold both authors to account.

Eysenck had an increasingly untenable position to protect, one that minimised the health effects of tobacco. His heterodox position had come with a long and lucrative funding stream from the tobacco industry. Grossarth-Maticek’s data enabled Eysenck to combat the criticism his position invited.

Defending entrenched positions, no matter what the cost to his credibility, was a feature of the latter part of Eysenck’s career. Two years before his death in 1997, Eysenck gave an intriguing insight as to what this might involve. Explaining why scientists might cheat, an apologia for his mentor Cyril Burt, Eysenck wrote:

"their finest and most original discoveries are rejected by the vulgar mediocrities filling the ranks of orthodoxy … The figures do not quite fit, so why not fudge them a little bit to confound the infidels and unbelievers? Usually, the genius is right, of course (if he were not, we should not regard him as a genius), and we may in retrospect excuse the childish games, but clearly this cannot be regarded as a licence for non-geniuses to foist their absurd beliefs on us. Such, then, are the conditions in which leading scientists may be led to falsify their data, or invent them. Convinced (rightly as it happens) that they have made a stupendous discovery, … they see that discovery threatened by enemies or hostile people with the power to destroy them. The first findings in trying to substantiate a theoretical discovery are never clear cut … But enemies would seize upon these anomalies to destroy his theory. Obviously the way to overcome this problem is simply to make sure the data fit the theory!
[4]

The problems posed by Grossarth-Maticek’s publications with Eysenck can be partly attributed to the considerable shortcomings in the research approval and review process at the time. Grossarth-Maticek’s research was an essentially independent undertaking: a long-running longitudinal prospective program, replete with a variety of interventions. His Heidelberg program was particularly extensive, involving up to 20,000 subjects. It dated back to the early 1970s, while his work in Crvenka began even earlier. As Pelosi noted, there were no written protocols for the Heidelberg programme, since it began at a time when they were not mandatory. Moreover, studies of interventions made during 1970s and 1980s never went through an ethics approval process in Germany, for the same reason. Despite being an Emeritus Professor at the IoP, Eysenck’s involvement did not trigger such a process in Britain either. The path to publication for their joint work largely avoided a thorough peer-review process. Much of it was published in journals Eysenck founded, or in what were pay-for-page journals, or in edited volumes overseen by sympathetic allies. Much of the scrutiny that has been possible has been post-hoc, post publication – the Psychological Inquiry symposium and follow-up articles are the key examples of this. In effect, this process is still going on. And it needs to be brought to a head.

While this problematic research remains in the literature it has the power to mislead. Avoiding digesting the ideas, citing the results, or including them in meta-analyses requires knowledge of the published critical articles. And these critical articles have only been able to address some of the many doubtful aspects of this work. This is not good enough.

Some of Eysenck’s defenders told me that replication efforts would cast ultimate judgement on the value of this work. Genuine attempts have been made, without yielding anything close to the spectacular results of the original work. In any case, the instructive value of replicating this work was limited by the lack of openness and detail from Grossarth-Maticek in particular.

Eysenck’s work with Grossarth-Maticek needs to be impartially and rigorously investigated. Grossarth-Maticek still has the data, to the best of my knowledge. In an era in which science finds itself in a crisis of trust, those called upon to conduct an inquiry should aim to do better than defer responsibility. If the conclusions of such an inquiry warrant it, more than 61 suspect publications of Eysenck and Grossarth-Maticek should to be flagged as untrustworthy or retracted.

Eysenck remains an influential disciplinary figure. It is time to set an example: to ensure the integrity of the literature, to ensure future generations of researchers don’t behave in the same way, and to ensure their colleagues don’t face the same dilemmas.

Those who truly believe in the scientific project need to do their duty.

References:

1 Pelosi AJ. Personality and fatal diseases: revisiting a scientific scandal. J Health Psychol 2019; 24: 421-39. 10.1177/1359105318822045 30791726; Marks DF. The Hans Eysenck affair: time to correct the scientific record. J Health Psychol, 2019; 24: 409-20. 10.1177/1359105318820931 30791728
2 Buchanan R. Playing with fire. The controversial career of Hans J Eysenck. Oxford University Press, 2010.
3 Eysenck, H.J. Reply to criticisms of the Grossarth-Maticek studies. Psychological Inquiry 1991; 2: 297–323.
4 Eysenck, H.J. Burt and hero and anti-hero: a Greek tragedy, in Burt: fraud or framed? (ed. Nicholas J. Mackintosh), pp. 111–29, Oxford University Press, Oxford, 1995, p.126.