Editor's note: Analysis of papers published in <i>JAN</i> in 2002

As part of our continuous quality monitoring activities for JAN, I have analysed the 256 papers published throughout 2002. This was a fascinating exercise and identified many issues for us in developing the journal, and this analysis was discussed at our Management Team Meeting and at the meeting of the International Editorial Board which took place in September 2003. In this editor's note, I will report the analysis and then draw some lessons for the future, not only for the editorial and publishing team, but also for JAN's worldwide network of reviewers, and future contributors of papers. In 2002, 766 papers were submitted to JAN from 35 different countries on all five continents. In general, the acceptance rate was approximately one third of the papers submitted from each country. Table 1 shows that the UK was the largest source of accepted papers (n = 101), followed by the USA (29), Australia (28) and Canada (21) as other substantial contributors of accepted papers. Seven European countries other than the UK contributed 41 papers, and China, taken as a whole (including Taiwan and Hong Kong), was another of the larger contributors (23). Thus, as anticipated, countries where English is the first language or the major language used in education were the source of most papers. However, the relatively uniform acceptance rate across all countries shows that language was not the determinant of whether papers were accepted. Language use is a major editorial issue, and is not confined to papers whose authors do not have English as their first language. Poor syntax and style are common even among UK authors, and papers from North American and Australasia often need substantial editing into standard English. Many of these problems could be reduced if authors made more use of the ‘readability’ guidelines on the JAN website. Final editing at the point of acceptance is now done either by the Executive Editor or by a copy editor (who is also an experienced nurse researcher). Each paper takes a minimum of one hour to edit, but may take several hours when substantial language improvements are needed. In a few cases a paper is returned to the author when first submitted, with a request to seek help to improve the English before the paper is of a standard to be sent for review. Occasionally this is also necessary at the acceptance stage if the language in the paper is very technical. Clinical topics were the focus of the largest number of papers accepted (n = 118), with nursing education issues being the second most numerous (29), closely followed by methodological papers and theory or concept development (22), 10 of the latter being concept analyses (Table 2). Table 3 lists the methodology used in papers, where this was stated by authors. In many instances the methodology was not delineated explicitly, and could only be derived from use of a term such as ‘qualitative’. Only a minority of empirical papers explicitly identified the methodology used and the majority of these claimed some kind of qualitative research approach. The largest group (n = 27) simply stated that the work was qualitative, with smaller numbers specifying phenomenology (16) and ethnography (10), and only one describing the methodology as grounded theory. The latter was stated as a ‘design’ in a further three papers (see below). Twenty-two were discussion or position papers. This leaves 108 JAN papers in which the methodology was not stated and/or could not be identified from the information given. Clearly, it is essential that methodology is properly identified and this is an issue now under closer editorial scrutiny. Again, only a minority of empirical papers explicitly identified the design that was used (n = 111) and, in many examples listed in Table 4, the type of design was identified only by terms such as ‘exploratory’ or ‘descriptive’ within the methods section of the paper. In some cases the term ‘design’ was not used appropriately, as for example with discourse analysis, fieldwork and grounded theory. Here, ‘methodology’ or ‘approach’ would be more suitable. The largest group of empirical papers (n = 49) were descriptive or exploratory in design. There was a total of 18 trials, of which eight were randomized controlled trials and 10 ‘cross-over controlled’ trials, with a further four quasi-experiments and two ‘experiments’. Three more studies were described as comparative. The lack of more sophisticated designs is perhaps disappointing at this stage in the development of nursing research. There is scope for more use of existing data in the form of documents, or secondary data analysis; and the need for replication studies and, arguably, more frequent use of economic analyses built into experimental and evaluative research. The data collection methods used in reports of empirical studies are shown in Table 5. Interviews (n = 50) and questionnaires (n = 45) were the principal methods reported as used alone, with a further 23 and five respectively using these in combination with other methods. Only 12 studies used physiological measures alone (n = 10) or in combination with others (n = 2). The principal methods used, and several of the less frequent ones, relied on self-reporting and this limitation is again disappointing at this stage in the development of nursing research. Only three studies using questionnaires employed standardized instruments, which leads to further questions about validity and reliability where authors designed their own questionnaires specifically for the study in question. Encouragingly, however, 51 studies used triangulation of methods, which should add to rigour. The majority of samples in empirical papers were of less than 100 (n = 116), with 25% being of 10 or fewer, as shown in Table 6. The type of sample was rarely stated, with six being population samples, one consecutive, one snowball and one purposive (although the ‘purpose’ of this was not given). The assumption must be, therefore, that virtually all the empirical studies were based on convenience samples. A justification for the sample size was not given in any paper, and no use of power calculations was reported. An adequate description of the sampling strategy and justification of the sample size are fundamental requirements in reporting research, and editors will now be playing closer attention to these. Table 7 shows that patients or clients (including lay carers) were a major category sampled either alone (n = 71) or together with nurses (13). However, healthcare staff and/or students made up the samples in 74 studies, as well as being included with patients/clients in 13. This preoccupation with studying nurses may reflect perceived difficulties in obtaining ethics committee approval for patient-focused research, and lack of funding for more ambitious projects. Whatever the reason, we are all aware of the greater importance that research funders now place on patient-focused studies that can provide evidence to underpin practice development and service improvement. JAN is a leading international nursing journal, and consistently gains the highest prestige rating among UK-published, peer-reviewed nursing journals. Clearly, no firm conclusions can be drawn about trends from a snapshot review such as this, but a lack of sophistication in designing and reporting evidence has been uncovered, and this needs to be addressed. Nursing has increasingly been a university-based discipline over the past two decades in the countries that are submitting to JAN. The nursing profession itself, and wider healthcare policymakers, have also increasingly called for practice to be evidence-based. Therefore it is disappointing to see that empirical studies reported in JAN still tend to be small-scale and descriptive, to use convenience samples and to rely heavily on self-report methods. Reviewers and editors cannot have any direct effect on the quality and scale of the research that is being conducted, but they do have an important part to play in influencing the quality of papers that are published, and this is certainly the case in JAN. Author guidelines and checklists for reviewers and editors have been developed, and these are available on JAN's website (http://www.journalofadvancednursing.com). On the other hand, there is a dilemma in being seen to be too prescriptive, especially when contributors may not have the resources to do more complex research or do not have the skills or support to report it more rigorously. However, because we want continually to drive up the quality of JAN, we have revised the guidelines for authors (also available at http://www.journalofadvancednursing.com) by simplifying them in some cases and adding more detailed recommendations in others. We hope that this will help authors to check that they have included the detail about empirical studies and literature review papers that is essential in any research report. Readers will make their own judgement about the quality of the work reported but, reasonably, they should be entitled to expect that the reports they read in JAN do contain all the essential details. These same guidelines are considered by reviewers during the appraisal process of papers submitted to the journal and also by the editors when making decisions about the need for revision of papers prior to acceptance and, finally, their publication. Our aim is that these transparent and public guidelines for authors and reviewers may help in setting standards for publishing in nursing by indicating what constitutes best practice in reporting research.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.