Accuracy of references in Journal of Cytology.

Sir,

References provide background material for the subject of article and support for the comments made or theories expanded and they demonstrate to the reader that the author has researched the topic in question comprehensively.[1] It is very important for the references to be correct, as incorrect references frustrate the reader while searching for related articles and may prevent the original data being identified. Despite editorial instructions for checking the references accurately before a manuscript is submitted for publication and the availability of various electronic resources, this continues to be a major problem in almost all the specialty journals. Previous authors have observed an error rate ranging from 3% to 60% in various general and specialty medical journals.[2] So far no such study is done regarding the accuracy of references in Journal of Cytology (JOC). So this study was done to observe the accuracy of references in the published articles of JOC.

Eighty nine references were randomly selected from all the four issues (1 to 4) of volume 27, 2010 of JOC. The instructions for authors of this journal were examined to know how to write the references. All the selected references were compared with the original for accuracy, using following six elements: 1) authors: Spelling, initials, extra/missing 2) title: Spelling, punctuation and missing 3) journal name for use of correct abbreviation as listed in Index Medicus 4) year 5) volume, and 6) page numbers: First and last page numbers. Citations were considered incorrect if there was an error in any of these six elements. If there were multiple errors in one reference, it was counted as one error only. References not cited from indexed journals were excluded. Text book references were also compared with original text book for accuracy. Statistical analysis was done by using frequency and percentage.

Out of 89 references, 82 were cited from the journals and 7 from text books. Out of 89 references 14 (15.7%) were inaccurate. Out of 14, 11 were from journal citation and 3 from text books. The most common error is author's name and was found to be incorrect in 8 (8.98%). Journal name, journal title/text book title, and page number error was identified in 2 (2.42%) each. There was no error with respect to Volume number and Year [Table 1]. Some of the examples of errors in citing the references are given in [Table 2].

Table 1

Frequency of reference errors in Journal of Cytology

Table 2

Examples of errors in references in Journal of Cytology

References serve as an important tool in providing credibility to the published literature and to assist retrieval of the cited and related information.[2] The international committee of medical journal editors states unambiguously that the references must be verified by the authors.[3] Citation of references with errors can make it difficult for the reader to retrieve reference and obtain, check or verify the information to which the text of the paper refers. Major errors such as volume, year, and page number make assess to the journal difficult and minor errors such as name of author impact negatively on the academic records of the author.[4] Various authors have identified different causes for the inaccuracy of references, the most common being carelessness of the author, over simplification of the results, information taken from book chapter or from online source without searching for the original source are some of the other causes.[1]

Roach et al.[5] observed an error rate of 60.7% in three major international obstetrics and gynecology journals, this is the high end of error rate reported. Adhikari.[1] observed an error rate of 30.1% in Indian Journal of Otolaryngology and Head and Neck Surgery. Vargas- Origel et al.[4] observed an error rate of 36% in Acta Paediatrica, 22% in Archives of Disease in Childhood, 29% in Journal of Pediatrics and 32% in Pediatrics. A study by Adhikari.[6] in the indexed journals of Nepal revealed an error rate of 11.6% in Nepal Medical College Journal, 11.1% in Journal of Nepal Medical Association and 23.3% in Kathmandu University Medical Journal. Our observation of error rate of 15.7% falls within the reported range. The error rate observed in other journals are 14% in Australasian Medical Journal,[2] 37% in National Dental Journals, 56.0% in Hong Kong Medical Journal, 27.5% in three major American Medical Journals, 37.5% in Otolaryngology/head and neck surgery.[1] The most common error observed in the literature is the author's name as observed in our study. In addition, we have also identified that the references cited from text books contain more errors than the one cited from journals.

In all the journals, in the instruction for authors section it is written that “authors are responsible for the accuracy of references and must verify them against the original documents”. But this is not strictly followed by the authors. The authors probably consider references are the least appendage of the manuscript and hence they fail to pay attention to it. Another reason for errors could be the misguided idea that no one (editorial team, reviewers, or readers) would care about the references listed. It is difficult by the editorial team to check all the articles submitted for publication, instead they can do a random checking of the references and if in any article, reference is found to be incorrect then complete checking of the references can be taken.[1]

Few of the journals including JOC may request the authors to send first page photocopy of cited article for cross checking of the reference. In journals using Journal on web for article submission there is a link for checking the accuracy of references, upto 15 references can be checked at a time. But this is restricted only to journals indexed by Pubmed.

There are many suggestions by the previous authors to decrease the error rate of references in the published article and some of them are a) directly consultation with the original work or obtain this work from a computerized data base, b) limiting the number of references, c) submitting the first page and last page of the referred articles, d) avoiding text book references, e) reviewer checking of a sample of reference from submitted articles, f) page proof sent to authors for final correction of the article before publication. Furthermore, the editorial board should make publishing requirements more stringent, because there is an association between the journal impact index and the number of errors.[14] So it is a team work by the authors, editors, and reviewers to decrease the error rate.