In this article we briefly describe the basic elements that form a scientific paper or a technical report. The situation we envision is the standard one: you have collected data and you must now document that data and interpret its meaning. For the purposes of this discussion, it is immaterial how the data were obtained: they may have come from laboratory experiment, from a computer simulation, or from a random survey. No matter how data are obtained, the reporting problems are basically the same. The elements of your presentation are these:
|1. Introduction||5. Summary|
|2. Methodology||6. References|
|3. Results||7. Appendices|
|4. Discussion||8. (and perhaps) Abstract|
The Introduction establishes a context for the work: what was done, why it was done, what was to be learned from it. This can often be done in one paragraph. If the problem has a history, then an additional two or three paragraphs might be needed to review that history. The review should summarize previous studies of the problem and state why those results are incomplete, inconclusive, or otherwise conflict with one another. Presumably, the present study is intended to resolve those ambiguities.
In this section you identify the data that were collected and describe how those data were obtained. Your emphasis here is on the raw data, not on quantities calculated from the data. The description of methodology should include the apparatus, analytic equipment, experimental protocols, calibration techniques, numbers and kinds of control groups used, verification procedures (such as repeating runs), quality controls, identification and reduction of statistical and systematic errors. Your objective is for the reader to understand what data were collected and to be able to assess the reliability of those data.
Here you describe how the raw data were analyzed to obtain meaningful results. This step typically involves calculations, such as statistical analyses or use of a mathematical model or both. You want to show how statistical and systematic errors in the raw data propagate through the calculations and into the computed results. Analyses often involve creation of tables, figures, and graphs to help identify trends, patterns, and correlations in the data.
The Discussion is the point at which you attach meaning to the data and the results. How do the trends, patterns, and correlations in the results help you address the problem you set out to solve? How do the data compare with theoretical predictions or existing mathematical models?
The technically unsophisticated often hold to the notion that attaching numbers to an activity makes it scientific. But numbers are only tools. Numbers combined with mathematical logic give us a powerful way to reason, but in the end, we still must interpret numbers and their implications.
The defining attributes of scientific activity are an objective assessment of alternatives and the elimination of those alternatives that are inappropriate, undesirable, or just plain wrong. Note we don't necessarily find the right answer; rather, we eliminate bad answers. Good experiments are always created in such a way that they must eliminate some of the possibilities. We then may have to do more experiments to further reduce the remaining list of alternatives.
Your Discussion should develop the logic that leads from the results to a conclusion that resolves your original problem. The Discussion should be quantitative; not A is better than B, but rather, A performs 20% better than B. You went to considerable trouble to obtain numbers; use them. The presentation should include not only why your solution is best, but also why alternative solutions are not as good. You must convince the reader that you have seriously considered all the alternatives. Moreover, you must state the limitations to your "best" solution and the conditions under which the solution is "best".
Here you summarize the contents of the previous four sections. Usually, just three or four paragraphs are required. A reader should be able to read only this Summary and learn what you did, why you did it, how you did it, what your results were, what those results mean, and what conclusions you deduce from that meaning.
This section should help a reader explore the background that is relevant to your problem. Thus, citations must be complete; they should include all authors, titles, editions, publication dates, and complete information about publishers. Use a format that is standard for your discipline: look at the format used in other documents, similar to yours.
The appendices should contain supplemental information that more fully documents your presentation but which is not necessary to support your discussion and conclusions. The main body of the report must be able to stand on its own.
Long reports or highly technical ones are often preceded by an Abstract—a four-to-ten sentence synopsis of the paper. The emphasis is on identifying the problem and stating the conclusions. If the methodology is unusual or otherwise particularly important, then one sentence of the abstract could be used to state the methodology.
Good abstracts are difficult to construct. A first draft can be made from the Summary that you prepared in Section 5, so it is better to work on the abstract after Sections 1-5 are drafted, even though the Abstract appears first and is the first thing encountered by the reader.
So that's a thumbnail sketch of what should appear in a technical report, but we've said nothing about how you achieve a coherent, cogent, concise report. In most endeavors, what you do is less important than how you do it. How do you combine sentences, paragraphs, tables, figures, and equations into a meaningful technical story? To learn how, you need the little book Technical Style—it's available from this website.