Wednesday, January 9, 2008

Communicating biogeography

Robert J. Whittaker was sole editor and then editor-in-chief of Global Ecology and Biogeography from 1995 to April 2004. He has been editor-in-chief of Journal of Biogeography since October 2004.

Some months ago, Joaquín Hortal approached me to write a piece for this newsletter, with the remit of providing some pointers as to what we look for in papers submitted to Journal of Biogeography. In some respects this is easy to do in that the journal policy is to be inclusive of major traditions within the field and international in scope, within which we give preference to papers posing and answering clearly phrased questions or testing hypotheses of general interest. However, having been given the opportunity, I also wanted to make some comment pertaining to presentation and rigour of papers in the natural sciences generally (i.e. not just submissions to Journal of Biogeography), which is a rather more risky thing to do. Not least because I might see my own words thrown back at me by referees of my papers in the future with the admonishment that I cannot follow my own guidelines as to best practice! I say this because I think it is deceptively difficult to write scientific papers well.
The process of writing the paper should not be regarded as just the bit that comes at the end, after all the challenging scientific stuff of laboratory analysis, experimentation and sophisticated statistical analysis has been done; rather it is integral to shaping the interpretation and developing ideas and arguments, which rarely come to the page fully formed. It involves establishing the context of the work through making multiple connections to the existing literature, which continues to grow at a phenomenal rate. All of this presents significant challenges to the scientific author, requiring a systematic and thorough approach to the task of constructing a paper. The end result must be accurate, meaningful and clearly communicated so that it is understandable by the target audience.

So how well are we doing and are there particular problems facing those starting out in biogeography? If this were a school report, I think the answer to the first part of the question would probably be an enigmatic ‘could do better.’ In illustration of concerns that I share, Todd et al. (2007) provide some quantification of the extent of mis-citations in a sample of 306 ecological papers selected from 51 journals, focusing on specific assertions made in the sampled papers. They report that some 7% of citations did not support the original statement at all, 11% were ambiguous, 6% were ‘empty’ (i.e. citing an opinion from a secondary source), with the remaining three quarters being ‘good’ citations. Todd et al. (2007) provide some comparative data suggesting that ecology may not be particularly unusual in having a significant level of sloppy citations, but, pulling no punches, they go on to conclude (p. 1660) that “mis-citing of references is widespread in ecology and is currently part of the normative values [my italics] of workers within the discipline.” I have given this example because: a) an effort has been made to quantify one key indicator of the rigour of scientific papers in a closely-related (overlapping) discipline to biogeography, and b) because it is something that we should all be capable of getting right. Based on my experience as a reader, reviewer and editor, I suspect that an analysis using the term ‘biogeography’ in place of ‘ecology’ would yield similar findings.

The question arises: how do errors arise within papers and what steps can we all take to minimize them? Todd et al. (2007) focus on the second part of this question, arguing that journals should ramp up their efforts to detect failings such as poor citation practice, implying that editors and reviewers should do more, whilst conceding that “…it is as authors that we have the most potential to improve citation practices.” My own take on this is that the onus should rest firmly and squarely on the authors of paper and not just with respect to citations but in all aspects of paper preparation and presentation.

One of my most rewarding roles as editor-in-chief of Journal of Biogeography is to scrutinise the final revised versions of all papers prior to formal acceptance. Whilst I get to read many wonderful and fascinating papers, setting out analyses of complex data sets with great skill, it is surprising how many papers (supposedly in their final form) contain significant flaws, internal inconsistencies, failures to follow guidelines, and failures to describe accurately the results presented in the tables and figures. Moreover, returning to citations, few papers (my own submissions to other journals included!) manage to compile reference lists that accurately match the list of papers cited. Such failings are despite the very best efforts of reviewers and editors and, as Todd et al. (2007) point out, they frequently sneak through into published papers in a wide range of academic journals. I suspect there are many reasons for such failures. These may include:
  1. Pressure to publish, get the next research grant, etc, leading to rushed work.
  2. Lack of guidance from senior colleagues or supervisors for those starting out in academic scholarship.
  3. On multi-authored papers, senior colleagues may leave it to a single, often junior, author to undertake the revision, failing to provide adequate input (or any input!) on subsequent versions submitted to a journal.
  4. On multi-authored papers, errors and mismatches may be introduced through sequential edits by different authors as they circulate drafts around by email (indeed I suspect that the error rate in multi-authored papers is generally higher than in single-author papers).
  5. Minor errors are introduced during revision as a by-product of responding to reviewer/editor requests to add or correct analysis, or to drop analyses, tables or figures. The complexity of content of the average paper today means that it is almost inevitable that authors fail to update all of the detail in the text of second or third versions of a paper.
  6. Some authors reason that if a submission to journal X has a 4 in 5 chance of rejection then it is not worth putting too much effort into getting things exactly right at first submission until they have seen if the referees like it, thus normalising a rather laissez-faire attitude to accuracy.
  7. The problems of writing in a foreign language (which applies to many workers in biogeography).
  8. The autocorrect function in the word-processor, combined with an inability to type accurately in the first place, introduces all sorts of minor errors, especially into reference lists, where they are hard to spot. So, some advice on Autocorrect: turn it off and instead run a separate spell check.
  9. The complexity of much analytical work in biogeography is such that it is only possible with modern computers and spatial statistics, meaning that it can be extremely difficult for authors, reviewers and editors to check whether results are reasonable. However, the frequency with which minor errors creep through to late stages of manuscript submission raises the suspicion that insufficient efforts are made by many authors to check that their results are correct by replicating/cross-checking their analyses.
  10. Finally, if the paper does make it through to production by a journal, there is always the possibility of errors being introduced during copy-editing or typesetting. Unfortunately, because everything is supplied electronically these days, many authors seem to think there is hardly any need to check their proofs. Think again!
I am sure that the above is a far from exhaustive list of ways in which errors become entrained in manuscripts during the process of their development, and I am also certain that they are not specific to biogeography. However, biogeography is a particularly diverse discipline, with many technical terms, sub-disciplines, and schools of thought, and it is a truly international subject. Whilst we can celebrate this diversity, some have also despaired of the failure of different schools of thought to engage fully within one another. I suspect this has at least something to do with deficiencies in the way ideas are developed and expressed in many biogeographical papers. So, I see it as very important that papers published in Journal of Biogeography, in addition to being accurate in their content and focused on a clear question or hypothesis of general interest, are also clearly written and as accessible as possible to a wide audience. At present, as Ladle (2008) reports, biogeography as a discipline is largely cryptic in the public sphere of the news media and the blogosphere, and in the UK at least, in school curricula. So, once you have overcome the “deceptively difficult” task of completing a paper, there is a further task, often neglected, of making steps to disseminate the findings in other media (e.g. through press releases) to help build a wider base of public understanding and appreciation of biogeography.

In summary, I concur with Todd et al. (2007) that the dash to publish is part of a complex array of factors leading to a degree of slippage from the rigour necessary in scientific work. In my view, the onus falls on senior academics, whether as colleagues, mentors, co-authors, supervisors, or educators, to emphasise repeatedly the importance of theoretical and empirical rigour in all aspects of science, from conception of a study to the preparation of each draft of a paper submitted to a journal. These comments may seem to suggest that the problems stem largely from inexperienced authors, but in fact I have seen well constructed and badly constructed papers written by people of all career stages, suggesting that much of it comes down to aptitude (i.e. traits of the authors) and training at the outset of a scientific career. In biogeography, there is no doubt that we have an increasingly sophisticated global community engaged in our science, as readily seen from the diversity of authors publishing in journals such as Journal of Biogeography. This provides evidence of a healthy and vibrant field. So, my critical comments should be seen in this light: as a community of scholars we are ‘doing well but could do better’ (so, still enigmatic but with a tone of encouragement). Finally, in purely pragmatic terms, if the goal is to get your paper published, it is worth following the sentiment of the old saying that if you watch the pennies, the pounds look after themselves. If you can prepare a manuscript in which all the fine detail is correct and clear, it is more likely that you will convince your target audience (reviewers, editors, and subsequent readers) that your findings are valid and meaningful, or (let’s not be too hasty here!) at least worthy of serious consideration and debate.

Preparing a manuscript for submission – some suggestions
  • Read and follow the journal’s instructions for authors
  • Make and complete a checklist of all the components of the paper that require cross-checking prior to submission
  • In attributing ideas, check primary sources
  • Not read it? Don’t cite it
  • Co-authors should read and check the final (or penultimate) draft of both original and revised manuscripts prior to submission
  • Ensure tables, figures and appendices are cited in numerical order
  • Check the Results and Discussion sections to ensure the content accurately tallies with the final content of tables and figures
  • Make sure you have your word processor set to the right dictionary (e.g. UK English or US English)

Ladle, R.J. (2008) Catching fairies and the public representation of biogeography. Journal of Biogeography, in press.
Todd, P.A,, Yeo, D.C.J., Li, D. & Ladle, R.J. (2007) Citing practices in ecology: can we believe our own words? Oikos, 116, 1599-1601.

Editor’s note: With this first commentary from Robert J. Whittaker we start a new series for the IBS Newsletter. We hope to open a discussion within the biogeographical community, starting with views from the editors of key journals publishing biogeography papers about current trends in publishing biogeography, the kind of manuscripts they would like to receive to be considered for publication in their respective journals, and how to improve the quality of our work. Here, the editors will have the opportunity to write on the topics of their choice, with the general aim of improving how research in biogeography is communicated to the scientific community and the public.

No comments: