Scientific papers generally follow a pattern. It is important that the style of writing is clear and appropriate. Equally, it is important that the argument follows a logical structure, from the introduction to the conclusions. Papers are sometimes rejected because the message is not clear. A logical and clear structure will maximize the chances of people understanding the significance of your work. What follows is a collection of various bits of advice that have been provided to authors, researchers and students over the years. The italic text in brackets helps to clarify what the various parts of a paper are for, which helps authors to figure out where each part of their argument and evidence belong. Of course, this structure only applies to empirical papers. While these form the majority of papers submitted to this journal, we also welcome other kinds of paper. Thus, there are other structures that may apply.
Abstract (A concise summary of the whole paper): You should probably write this last. It would be difficult to summarize something until it has been written. For the sake of clarity and consistency within this journal, we ask all authors to avoid phrases that begin with “this paper…”, “this research…”, “this study…”, etc, especially in the abstract and conclusions. For guidance on writing an abstract that is consistent with the journal’s style, please refer to http://www.tandf.co.uk/journals/authors/rcme_abstracts.asp.
Introduction (What is it about?): The opening section of the paper should introduce the topic, and explain the kind of problem that is being dealt with in the paper, in general terms. Introductions often conclude with a section about the structure of the paper, although this is often unnecessary because a quick glance at the main headings will satisfy anyone who wants to know how the paper is structured. Indeed, many research papers follow a pretty standard structure, in which case there is no need to explain it. If you do choose to describe the structure of the paper, it is important to do more than merely describe the headings. A better approach is to explain how you have structured your argument and the major steps in your research. The purpose of describing the structure of the paper is to make clear the argument, rather than describing the headings. For more details on writing an introduction, see http://bit.ly/CMEintro
Literature review (What do we already know? About the specific problem. About problems of this kind and metaphors. What is it like? What is it not like?): A literature review needs good structure. A good literature review starts with one or two major (early) authors and explains how their thinking has inspired other researchers. There should be clear logical links and various strands of thought should be clearly identified. Further advice is available at http://wp.me/p1J7za-3h
Problem statement/research question (What do we need to discover? What kind of question is this?): This is the pivot point for the reporting of research. Up to this point, the paper is focused on what is known. From this point forward, the focus is the discoveries and how they were made. The research question should be clear and specific. It should be stated in terms that make clear that it is achievable. And it should form a logical conclusion to the review of literature.
Methodology (How are these issues usually researched, generally, and why? Another kind of literature review): The way that a question is phrased implies something about the kind of research that it will entail. Many papers are submitted without a methodology section. Many use the word methodology when they mean method. This is a perennial problem in every kind of scientific communication. Method refers to what you did, whereas methodology is the science of the method, which explains why the selected methods are appropriate for generating the kind of data and/or information that this kind of question demands. While it may appear to be pedantic to insist on these specific word usages, what is crucial is that the researcher makes clear why the chosen methods suit the question.
Method (What kind of data will be collected specifically; how will it be collected and how will it be analysed?): After a clear exposition of the methodological issues, the method will simply explain what was done. The particular techniques that were used, the controls that were put in place to ensure validity, the dimensions of the research strategy and so on, all need to be explained.
Data, analysis, discussion (What have we discovered?): A clear explanation of the method sets the scene for the data to be reported and summarized. It may not be necessary to include all of the data in its raw form. But on-line repositories can be used for archiving research data in the public domain. Such stores can be cited in case readers wish to follow up on the precise data. The presentation of the data should be separate from the reporting of the analysis. Similarly, the discussion should be restricted to discussing what the data has revealed.
Conclusions, limitations, recommendations (What does it all mean? So what? Who cares?): Perhaps the most important part of the paper is the conclusion. This explains what has been proved, and should also explain the parameters that make clear how generalizable the findings are. The limitations of the work are not intended to be agonizing self-criticism, merely the parameters of the data and therefore of the findings. For example, the study was undertaken at a particular time, in a particular country, profession or sector, thus the findings can only be generalized to a certain extent – the limitations explain the extent to which generalizations can reasonably be drawn. Recommendations may be directed at future researchers, they may be directed at practitioners or they may be directed at policy-makers. Any one of these three groups is sufficient. It would be unusual to be targeting all three.
Overall, the paper should begin with general things and, as it progresses to the research question, focus on to specifics. However, at the end, you should be linking the findings back to the general points and explaining the extent to which generalizations flow from the work.