Sunday 23 September 2007

Medical Hypotheses Impact Factor 1.3

Charlton BG. Medical Hypotheses 2006 impact factor rises to 1.3 – A vindication of the ‘editorial review’ system for revolutionary science. Medical Hypotheses
2007; 69: 967-969

Editorial

Summary

The Thomson Scientific Impact Factor (IF) for Medical Hypotheses has risen to 1.299 for 2006. This means that the IF has more than doubled since 2004, when it stood at 0.607. Using Elsevier’s Scopus database; in 2004 there were 437 citations to Medical Hypotheses papers published in the previous two years – by 2006 this had trebled to 1216 citations. Monthly internet usage of Medical Hypotheses run at an average of about 26 000 papers downloaded per month. An IF of 1.3 means that Medical Hypotheses has now entered the mainstream level of ‘respectable’ medical journals, in terms of its usage by other scientists. This is particularly pleasing given the aim of the journal is to publish radical and speculative ideas. A healthy IF is important to Medical Hypotheses because the journal deploys a system of editorial review, rather than peer review, for evaluation and selection of papers. Editorial review involves selection of a journal’s content primarily by an editor who has broad experience and competence in the field, assisted by a relatively small editorial advisory board. The great advantage of editorial review is that it is able, by policy, to favour the publication of revolutionary science. But since editorial review relies on hard-to-quantify and non-transparent individual judgments, it is important for its outcomes to be open to objective evaluations. Scientometric measures of usage such as citations, impact factors and downloads constitute objective evidence concerning a journal’s usefulness. Since Medical Hypotheses is performing adequately by such criteria, this provides a powerful answer to those who fetishize peer review and regard any other system of evaluation as suspect. Journal review procedures are merely a means to the end, and the end is a journal that serves a useful function in the dynamic process of science. Medical Hypotheses can now claim to perform such a role.

***

I am pleased to report that the Thomson Scientific Impact Factor (IF) for Medical Hypotheses has risen to 1.299 for 2006. This means that the IF has more than doubled since 2004, when it stood at 0.607 (www.scientific.thomson.com).

The IF is (approximately) a measure of the average number of times a paper in a journal is likely to be cited. Although there are important differences in citations according to research fields, and although IF has limitations if used to evaluate the potential importance of specific articles or specific scientists over the short term; nonetheless, I regard the general level of IF as a broadly valid measure of a journal’s importance among scientific peers.

Another measure of a journal’s profile in the scientific literature is the total number of citations per year, and here too Medical Hypotheses is thriving. Using Elsevier’s Scopus database (www.scopus.com), in 2004 there were 437 citations to Medical Hypotheses papers published in the previous two years – by 2006 this had trebled to 1216 citations. The journal’s influence is clearly expanding.

Furthermore, the monthly internet usage of Medical Hypotheses 2005–6 runs at an average of about 26 000 papers downloaded per month, which again indicates a very healthy level of interest from the broad scientific community (www.intl.elsevierhealth.com//journals/MeHy).

Aside from professional scientific considerations, Medical Hypotheses has an unusually high media impact (as can be seen by looking at internet news sources or performing web searches). Of particular interest to Medical Hypotheses readers is the imminent publication of a book about the journal, written by Roger Dobson and published by Cyan Books (London, UK). The book is titled: Death can be cured: and 99 other Medical Hypotheses. I have written a foreword, and can recommend the book as an edifying and amusing journey through some of the more stimulating ideas that have been published in the journal over recent years.

An IF of 1.3 means that Medical Hypotheses has now entered the mainstream level of ‘respectable’ medical journals, in terms of its usage by other scientists, and the probability is that this figure will rise further over the next three years. This is particularly pleasing given the aim of the journal is to publish radical and exploratory ideas [1] and [2], which inevitably means a greater risk that papers will be ignored by other researchers as being too speculative.

Another reason that a healthy IF is important to Medical Hypotheses is that the journal usually deploys a system of editorial review, rather than peer review, for evaluation and selection of papers.

Peer review in journals is used in a wide range of academic subjects including the arts and social sciences, and for most of its history science did not use modern methods of peer review; so that contrary to common assertions, peer review is neither distinctive nor essential to the natural sciences. What characterizes science is in fact evaluation by ‘peer usage’ – extrapolating facts and ideas to predict future observations, and making such observations in order to test published facts and ideas [3] and [4]. In other words, scientific evaluation comes after publication, and not before.

Currently, peer review in journals is a system whereby a prospective paper is vetted by gathering the opinions of two or three specialists in the field, any of which can (in practice) usually veto publication. The result of rigorous peer review (although most peer review, being unpaid and done as a favour, is probably far from rigorous) tends to be that published work is more reliable but less ambitious [1] and [5]: peer reviewed publications tend to be incremental advances on previous knowledge, performed by accredited professionals, and attained by standard techniques [4].

If Kuhn’s distinction between ‘revolutionary’ and ‘normal’ science is used [6] then peer reviewed research is therefore usually valuable for normal science, and when rigorous may then enable research to be regarded as probably valid at the time of publication. But the intrinsic tendency is for peer reviewed research to be limited in its scope and ambition so peer review intrinsically discriminates against revolutionary science [5].

Since Medical Hypotheses aspires to be a journal of revolutionary science ideas [1] and [2], it is clear that this aim would tend to be thwarted by peer review. Research which aspires to be ‘revolutionary’ science is relatively unlikely to get through the innate conservatism of peer review [5], since bold ideas are indeed less likely to be correct than cautious ideas. Looking back at the golden age of revolutionary biological science during the mid-twentieth century, it seems obvious that modern peer review would likely have killed the necessary ferment of radical speculations; some which turned-out to be accurate, but most of which were mistaken [7].

Until a few decades ago, the evaluation of scientific papers for journals was mainly by a process which I term ‘editorial review’. Editorial review involves selection of a journal’s content primarily by an editor who has broad experience and competence in the field, assisted (to a greater or lesser extent) by a relatively small editorial advisory board. The great advantage of editorial review – from the perspective of Medical Hypotheses – is that it is able, by policy, to favour the publication of revolutionary science.

Naturally, editorial review relies on hard-to-quantify and non-transparent individual judgments, which is why it is particularly important for its outcomes to be open to objective evaluations. Scientometric measures of usage such as citations, impact factors and downloads – albeit incomplete and imperfect – are all objective evaluations which tend to quantify a journal’s usefulness. So long as Medical Hypotheses is performing adequately by such criteria, this provides a powerful answer to those who fetishize peer review and regard any other system of evaluation as suspect.

In the end, journal review procedures are merely a means to the end, and the end is a journal that serves a useful function in the dynamic process of science [3]. Medical Hypotheses can now claim to perform such a role.

Acknowledgement

Thanks are due to Tanya Wheatley, Senior Publishing Editor in Health Sciences at Elsevier, for some of this data; and for her invaluable editorial support and strategic advice. I also like to thank the authors of Medical Hypotheses publications during the time of my editorship so far: upon their work, everything else depends.

References

[1] D.F. Horrobin, Ideas in biomedical science: reasons for the foundation of Medical Hypotheses, Med Hypotheses 1 (1975), pp. 1–2.

[2] B.G. Charlton, Inaugural editorial, Med Hypotheses 62 (2004), pp. 1–2.

[3] D.L. Hull, Science as a process, Chicago University Press, Chicago (1987).

[4] B.G. Charlton, Conflicts of interest in medical science: peer usage, peer review and ’CoI consultancy’ (Editorial), Med Hypotheses 63 (2004), pp. 181–186.

[5] D.F. Horrobin, The philosophical basis of peer review and the suppression of innovation, JAMA 263 (1990), pp. 1438–1441.

[6] T.S. Kuhn, The structure of scientific revolutions, Chicago University Press, Chicago (1970).

[7] H.F. Judson, The eighth day of creation: makers of the revolution in biology, Jonathan Cape, London (1979).