Main

Authors wishing to see their research findings published now have a much greater choice due to the vast increase in the number of medical and dental journals over the past ten years. Yet, despite the importance of publishing research findings, there are few reports in the scientific literature on the factors taken into account by editors in deciding to accept or reject a scientific paper. Currently, the only guidance provided for authors can be found in each journal's detailed instructions on house style and guidelines for authors. If further information was available and there was a specific pattern of requirements for publishing in dental journals, this would considerably assist authors in submitting papers that met the requirements of dental journals. Further, it would help editors to speed up the refereeing and correction phase of publishing.

The aims of this study were 1) To establish the factors that editors considered most important in manuscripts submitted to their journal. 2) To investigate factors which authors could address to expedite publication.

Materials and methods

Fifty dental journals were chosen (Table 1) and a questionnaire (figure 1) posted to the editors with a request to return the questionnaire by 4 weeks after the initial posting. The editors were informed that the results would be confidential. The questionnaire was piloted on 5 research workers with refereeing experience prior to being posted to the selected journal editors.

Table 1 Table 1
Full size table
Figure 1
figure 1

The questionnaire

Full size image

The journals were selected using the following criteria: 33 were identified from the list of journals indexed for dentistry in the Index Medicus (National Library of Medicine, Bethesda, Maryland). Review journals were excluded as well as those journals not publishing in the English language. Six further journals were identified from the SCI Citation Reports for Dentistry and Odontology. The remainder of the journals were chosen from areas of dentistry under-represented by the two previous listings. Eight journals were retrospectively regarded as 'generalist' and separate analyses were performed with and without responses from these editors.

The questionnaire occupied two sides of a single A4 sheet. A follow up mailing was sent after 6 weeks. Mann–Whitney tests with appropriate adjustment for multiple comparisons were used to test for differences between ranks in response to Question 1. A 'p value' less than 0.01 was regarded as statistically significant.

Results

Forty two editors responded, representing an 84% response. One editor did not complete the first question as he considered it was not an appropriate research tool, as he wished to rank some factors with equal weight. Two other editors commented on finding a similar difficulty with Question 1, but were able to complete the questionnaire. 6 replies were received from journal editors of the 8 journals which were retrospectively regarded as 'generalist'. Separate analysis, without responses from editors of these journals, gave a similar pattern of results to analysis which included responses from all journals. The results of question 1 (Table 2) are presented for the 35 specialist journals only.

Table 2 Table 2
Full size table

All editors except one felt able to answer the additional questions about other factors influencing the acceptance or rejection of a manuscript and the editorial policy on referees. The open questions produced the following responses. In response to the question 'What other factors might influence your decision to accept or reject a manuscript?', several editors mentioned 'poor construction of the paper' (cited by 20 [49%] of respondents) and 'poor research design' (15 respondents, 37%). A factor which editors valued highly was 'scientific novelty and timeliness of the topic' (12 respondents, 29%). Factors that caused most problems were 'poor use of English and careless preparation of the manuscript' (19 respondents, 46%). To expedite publication 'attention to guidelines to authors' was cited by 28 (68%) editors. Another factor adversely affecting publication was the duplication of that topic with similar results in recent publications of their journal.

Of the editors who commented on the use of the referees reports, one stated that most 'heartache' was caused when referees disagreed. Only eight (19%) editors specifically mentioned the use of the referee's report in judging a manuscript. Other factors that editors disliked were either 'unnecessary length' or 'over referencing of a paper' and three editors denigrated 'salami publishing'. 'Salami publishing' is a term that is used for multiple publications presenting results derived from a single experimental protocol, often from a thesis. One editor also commented that publications derived from a thesis often led to undesirable papers that were over-argued and boring.

All respondents except one answered the questions on their journal's refereeing policy. The majority of journals (28 respondents, 68%) used two referees, the remainder used three, except one journal which used four referees. Pro-formas to assist with the refereeing process were used by 36 (88%) of the journals. Over half (22 respondents, 54%) of the journals attempted to maintain anonymity of the authors.

Table 3 shows a distribution of the use of statisticians to review papers with the majority of journals using a statistician for less than 20% of submitted manuscripts, with three journals never using a statistician at all. Eleven percent of journals stated they used a statistician for over 60% of submitted manuscripts.

Table 3 Table 3
Full size table

Only five (12%) journals always circulated referees' reports although 18 journals (44%) occasionally did so. Some editors commented that this was done only when there was a disagreement about the merits of the paper or referees requested information. Two editors stated that routine circulation of referees' reports was a recent innovation.

Discussion

For authors of scientific research there can be no worse fate than rejection for publication, unless the science itself is flawed. Rejection because of lack of attention to detail or lack of awareness of the publication requirements of editors is also extremely frustrating and can waste a great deal of time as authors struggle to satisfy the differing (and sometimes inexplicable) demands of journal editors.

However, most editors of scientific journals are not just looking for papers to publish but are seeking papers which satisfy the requirements of their individual journal's profile. These requirements reflect the editorial policy of the publication itself and the particular philosophy of the editor, which may change when a new editor is appointed. It would seem sensible therefore for authors to know both the publication requirements and the personal editorial philosophy of the editor. While this seems obvious, experience shows that most authors pay scant attention to the 'notes for contributors' provided by the publication, let alone the personal philosophy of the editor. This is indirectly supported by a study that investigated author's criteria for selecting journals for submission of manuscripts.1 Respondents ranked the journal's prestige as the most important criteria, and the makeup of the readership and whether the journal publishes papers on the topic equal second. The existence of good editors, useful reviewers comments or advice on biostatistical methods were 7th, 8th and 11th respectively out of a list of 13 factors.1 The present study provides further evidence that supports the contention that authors ignore guidelines, in that attention to guidelines was mentioned by 69% of editors, who would hardly have done so if authors were already punctilious in this respect. The reply to question 1 suggested that attention to guidelines is one of these issues that editors will overlook if necessary, particularly if the work is of great significance and importance. However, obviously this will not endear the editor to the authors.

This study confirms that most editors are willing to provide information when asked, as evidenced by the high return rate of the questionnaires. Further, the care that editors took in considering their replies was encouraging demonstrating that they are keen to help authors expedite publication of their work.

The present study was partly initiated by previous work undertaken by Grace2 who sampled 51 delegates at the International Association for Dental Research (IADR) in San Francisco (March 1996) and 39 delegates at the British Society for Dental Research (BSDR) meeting in Bristol in April 1996. The questionnaires were provided for delegates to complete (on a self-selected basis) at an exhibition stand as part of an informal quiz. The questions Grace used formed the basis of the questionnaire for the present study. Delegates were asked to choose the two most important factors that editors took into account when deciding to accept or reject a manuscript. The results of these studies are shown in Table 4. The research workers who completed the questionnaire placed a greater emphasis on the 'relevance to the readership' than 'the content of the paper'. This was reversed in the present study with editors. The researchers also placed 'attention to guidelines' higher than 'spelling and grammar', and although the editors reversed the ranking of these factors, the difference was not statistically significant. Care has to be taken interpreting the work by Grace as this questionnaire was set up as a quiz and the results may be biased.

Table 4 Table 4
Full size table

Holt3 reported the results of a survey of 23 international and national educational journal editors by Noble which revealed that the appearance of the manuscript and the attention to guidelines for authors were the most important factors that editors took into account when assessing manuscripts. These findings are also in contrast to the present study where the dental editors considered scientific content and relevance of the work to the readership most important. One possible explanation for this difference is that Noble investigated journals publishing in the field of educational research, much of which is based on studies using qualitative methods, rather than the quantitative research more commonly seen in scientific dental journals.

Although the percentage of editors mentioning the referees' reports was relatively small 19%, we assume that editors took account of these reports in deciding that the paper was based on a 'poorly constructed research project' or the conclusions were 'poorly argued from the results of the study'. At present, 54% of editors state that they attempt to maintain anonymity of the authors from the reviewers. Anonymity is a complex issue as referees in a particular area will be aware of the work of other groups. However, studies have shown that blinded referees provide more unbiased reviews than non blinded reviewers.4,5 Abby, Massey, Galandiuk and Polk6 investigated the fate of rejected manuscripts as a tool to study the value of peer review. This retrospective study of rejected manuscripts in the Journal of Surgery over a one year period concluded that the review process acts as a sieve, influencing the standard of papers published in core journals, and should be considered as a valuable process. They did not comment on whether this was an indication of the effectiveness of the refereeing process or whether their findings could be generalised to other medical subject areas or non US publications. It is important that referees should remain objective and simply comment on the quality. The use of pro- formas would help this process, and 88% of the journals that replied regularly use them.

There was a wide range in the reported frequency of the use of statisticians as referees (Table 3). This may reflect variation in either the type of manuscripts which are published by each journal, many not requiring the services of a statistician, or the perceived ability of editors and their referees to comment on the statistical methods. However, the number of papers that have statistical referees was low and it will be of interest to speculate as to whether this will increase with more complex statistical methods being applied to dental research.

The data from the present study has revealed some important responses of a sample of editors of scientific dental journals. This data should be helpful for dental research workers in the preparation and submission of papers to journals.

Conclusion

The results presented here show that editors value scientific novelty highly, however poor construction of the paper and poor research design will lead to rejection. Manuscripts should be targeted at appropriate journals complying with the stated aims of the journal. Over two thirds of editors cited the importance of complying with the journal's guidelines as a means to speed the process of publication.