The word ‘study’ can be confusing because it has several meanings, inter alia aninvestigation of certain phenomena, and a written report of such an investigation. However, for simplicity, in this guide ‘a study’ always refers to an investigation, and a manuscript describing an investigation is referred to as ‘a paper’. Clearly, before starting to write a paper describing a study, it is essential to decide what the paper is going to cover, that is, the study must be delimited. Sometimes this is easy.
For instance, a study could be summarized as follows:
- Rationale: It is generally believed that mature bananas are yellow and bent. However, the Learned Society of Unorthodox Thinkers (a fictitious body) has recently postulated that they are in fact red and straight, and if they aren’t they certainly should be.
- Objectives: To test the general belief and the Learned Society’s conflicting hypothesis.
- What was done: Two thousand mature bananas were acquired and examined.
- Findings: All of the mature bananas examined were yellow and bent.
- Implications: The results indicate that bananas are generally yellow and bent, supporting the traditional belief (although it is possible that some are red and straight, since the survey was far from comprehensive). Whether they should be red and straight requires further investigation.
In this case, delimiting the study is very straightforward. It is also often straightforward in other cases where one or two simple hypotheses are postulated and tested. However, it is not generally quite so easy, because most investigations are much more complex. For example, in PhD projects multiple phenomena are often investigated, which could be reported in (say) three long papers, or larger numbers of short papers. Clearly, in such cases it is essential to decide which parts of the project are going to be covered in a particular paper. Fortunately, researchers usually have intimate knowledge of the scope (and linguistic style) of papers published in journals covering their fields of interest, and this can provide a good guide for deciding how much information should be included in each paper and thus dividing the project into a series of studies.
In addition, the elements of a larger project can usually be grouped into a set of reasonably discrete investigations, which greatly facilitates the delimitation of studies. For example, let us consider the following hypothetical project. The small, fictitious country Sucrosia has a near-ideal location and conditions for producing sugar from sugarcane, hence sugarcane is cultivated in large areas of the country, after which it is milled and the resulting sugar is refined for export. Some of the waste biomass (‘bagasse’) is also used for cogenerating energy. However, the yields and profitability are generally low by international standards. There are grounds for thinking that the poor yields are partly due to over-fertilization. Thus, this possibility clearly needs to be addressed, but many other variables also need to be considered, including the irrigation strategy applied, the cultivars used, harvesting operations
and scheduling, field-to-mill transport, cut-to-milling delays, the milling and other processing equipment, the use of steam and power, energy cogeneration options and waste treatment. In such cases, as illustrated in the figure above , flow charts may be very useful for visualizing the work to be done, delimiting studies within the project (and assigning human and other resources to them), tracking progress and (eventually) writing sections of papers and/or reports.
Sub-dividing a project in this manner can also provide a very convenient means for constructing sub-titles of sections of interim and final reports, by simply slightly re-wording the summarizing terms (shown in bold) for each of the delimited studies (e.g., Optimization of the fertilization regime, Optimization of the irrigation regime and Milling and processing strategies). Of course, each of these studies may be quite complex, so further. Such detailed flow charts can provide templates for writing sections of papers , and can be useful for identifying references that need to be cited in each section. It should also be noted that there will be substantial overlap between some sections of papers describing these studies, for instance, plots at the same study sites will probably be used for the fertilization, irrigation and cultivar selection studies, so the descriptions of the sites, and the criteria used to select them, will be the same in Studies A, B and C. This is convenient, because these aspects of the studies need to be described in detail only once, and after (say) writing a paper on optimization of the fertilization regime, papers on optimization of the irrigation regime and cultivar selection can refer to information in the first paper.
Having divided a project into discrete studies that can be described in papers of appropriate length, a related problem is deciding where to start from, that is, what aspect of each study to describe first. Some authors recommend starting by describing what was done, that is, the Materials & Methods section of a scientific paper in traditional format, or the findings, that is, the Results section (Malmsfors et al. 2004, Gustavii 2008, Booth 1993). Starting with what was done has some merits, since it is the only aspect that is certain (e.g., there may be uncertainty about what to include in an Introduction or Discussion, but provided good records have been kept, there should be little doubt aboutwhat was done). Similarly, the researchers will have clear knowledge about their results (although their implications may be disputed). Furthermore, having written either of these sections, the resulting text can provide a framework on which to base the rest of the paper, to ensure that all sections are consistent.
However, this raises two problems. First, it is essential to know exactly what to include in these sections, which can only be decided after delimiting the study. Second, it is essential to describe the Materials & Methods (and Results) in a logical order. Generally, the optimal order is chronological, for reasons discussed later (although other approaches to organizing material can be used, see Matthews and Matthews 2008). However, investigators might only remember to analyze certain control samples that should have been analyzed in early stages of an investigation toward its end. In such cases, they would seem foolish if they presented what they did in the true chronological order, stating at the end of the Materials & Methods section We then analyzed the controls, which we had previously forgotten to do. Instead, it would be far better to state at an earlier point that Both the extracts and controls were analyzed. Similarly, a substance that has taken months to purify may be dropped, scraped off the floor, re-purified and then analyzed. In such cases we would not recommend stating The substance was dropped, scraped off the floor, repurified and then analyzed. Instead, we would write, simply The purified substance was then analyzed. Thus, the Materials & Methods section should present what was done, or rather what would have been done if everything had been done correctly the first time in an ideal order, which may not coincide completely with the order in which everything mentioned was actually done.
For these reasons, a framework (which should be clear, simple and consistent) is required before starting to write this or any other section. Fortunately, such a framework can be constructed, for any study, by briefly describing the rationale, objective(s), what was done, the findings and the implications of the study. The way in which such a framework can be used to compose each section of a paper is described in detail in following parts of this guide, but before doing so we should define these terms, recognizing that a scientific investigation is rarely a smooth progression from an initial rationale, through formulation of a set of testable hypotheses, to experiments that have been perfectly designed and executed, yielding perfectly analyzed and interpreted results. Thus, here:
- Rationale refers to the context or background of the study, as understood at the time of writing, which may not fully coincide with the initial rationale. For instance, the initial rationale may have been partly based on a misunderstanding of a previous author’s work. If so, we would not recommend writing that certain hypotheses were tested because we misunderstood Smith’s conclusions, but instead adjust the rationale. Note, this is quite different from cases in which a well-grounded hypothesis formulated from a sound initial rationale was tested and found to be false, for which there is no need to adjust the rationale.
- Objectives refers to the specific goals of the study as understood (with hindsight) at the time of writing, which again may not fully coincide with the initial objectives, since the goals may shift during the course of the study; some may be added, some may be dropped and others may change. Thus, for instance, it would usually be pointless for an author to describe the context of a hypothesis that he/she initially planned to test, but did not because there was insufficient time, except perhaps in the conclusion, if possible future analyses are mentioned.
- What was done refers to the experiments and analyses that were performed, in the order that they should ideally have been performed.
- Findings refers to the results from those experiments and analyses, and the conclusions that can be directly drawn from them.
- Implications refers to conclusions that can be indirectly drawn or inferred from the findings, for example, whether a tested substance could be viably used in a proposed application, with referenced comparisons to previously published findings.
Initially, statements describing these aspects can be very short. Indeed, writing short statements describing each of the aspects is essential for composing key sections of a paper (especially the Title, Abstract and Conclusion). In addition, a fuller, much more detailed framework can be very useful for checking that all aspects of the paper are consistent, in other words that:
- the rationale provides sufficient context to justify everything that was done
- the description of what was done details all the materials used, treatments applied and experiments for which results will be mentioned
- results of all experiments mentioned in the what was done section are covered
- all of the main findings are discussed and
- appropriate references have been added at appropriate places.
How such a framework can be drafted and applied in practice are the main concerns of the rest of this guide.
Blackwell, J., & Martin, J. (2011). A scientific approach to scientific writing (p.6). Springer Science & Business Media.