Deskins Writing Tips

Professor Deskins' Writing Tips

Advice for writing MQPs, IQPs, theses, dissertations, and journal articles.

1. Write daily. Write often, even daily - at least 30 minutes per day. It is much better to spread out your writing than trying to cram it. Research even shows that writing daily is much more effective than sporadic writing. Keep a log of how much you have written to keep yourself honest.

You may have heard of writer's block. If you do things properly it shouldn't happen. You may feel stuck, but there are so many things you can do to advance your work. Create a figure. Make a table. Add some references. Write one paragraph.

2. Write paragraphs using topic and support sentences. A simple pattern for each paragraph is to have a topic sentence, followed by support sentences. Your topic sentence (usually near the beginning of the paragraph) summarizes the main point of the paragraph, while the support sentences provide evidence/arguments to support your topic sentence. A sample paragraph is below.

Carrots apparently have the highest amount of vitamin Q of any vegetable. Our gas-chromatography data, collected for fifteen common vegetables, is summarized in Figure 3. Figure 3 shows that carrots have a concentration of 15 mg vitamin Q/g vegetable, which is much higher than the second-highest vegetable, tomatoes. Tomatoes only have 5 mg vitamin Q/g vegetable according to our results. Furthermore Smith and Jones[1] found that carrots have on average 14 mg vitamin Q/g vegetable using high-performance liquid chromatography, in agreement with our results.

One easy way to help you in writing paragraphs is to write out a sentence for each paragraph what the main point of that paragraph is. E.g. "Describe simulation parameters", "Discuss differences of surfaces", "Discuss the linear regression model", "Compare our results with literature", etc. You can always delete this sentence, but it provides some structure/clarity for you when writing.

3. Be prepared to revise. Your first draft will likely need lots of revisions. In fact, your second, third, and fourth drafts may need lots of revisions. This is normal, including with my own writing. This is why you need to start early and often when writing! Don't be discouraged if your first draft isn't perfect. Revise it as necessary and move on.

4. Start with an outline. The easiest way to get frustrated writing is to just start writing without knowing what you're going to say. An outline makes writing so much easier since it provides a template and direction for what you should write. Start with the main document sections (Introduction, Methodology, Results/Discussion, etc.) as your main topics and then decide what sub-topics should include. Below is the start of a sample outline. After you have your outline, you can start to put a couple paragraphs in each sub-section. See this article for an excellent discussion on outlines.

I. Introduction
II. Background
 A. The Basics of Fuel Cells
 B. Limitations/Problems of Fuel Cells
 C. Alloys as Cathode Catalysts
 D. Molecular Modeling Techniques
III. Methodology

5. Make pretty pictures. People are visual creatures. It is much easier to understand experimental results when you have a plot to look at, compared to simply reading about the results. Also, understanding how a piece of equipment or process works is also easier with a picture compared to simply reading about it. The best scientific documents have good pictures. Of course, not all data looks good in a plot, so you may need to use tables. But even tables are easier to understand than reading data in a paragraph.

Make pictures and plots for all your important results and points. Your figures will serve as the anchors of your arguments. Once you have your figures in place, your writing will largely focus on explaining and interpreting them. Your figures are so important that after you have the outline (or while making the outline) you should decide which figures and tables you will include.

Warning! Do not use the Excel defaults for scientific plots. They are terrible. Label all the axes, including the units. The background color of plots should be white, not grey or some other color. Delete the grid lines and lines surrounding figure. You also don't need a title for the plot. That's what a caption is for. The caption should explain everything about the plot so the reader can understand it. For examples of some good figures see this paper. Also see this page.

6. Know your audience. Most of the people reading your report will not understand what you are talking about. They will likely be engineers and scientists, and may have heard of some of the terms (e.g. fuel cells or molecular modeling), but they will not really understand the details and what these mean.

Your job is to explain these things. Don’t assume your readers understand everything, just because you do. Explain, explain, explain. Especially specialized techniques, terms, or acronyms. Think about whether your fellow students would understand your report. Before you started the project, what would you have understood? Of course you don't have to explain everything since your readers will have some basic knowledge (your readers should have a basic understanding of chemistry, for instance), but you can't leave them lost. Give the reader enough details to understand your work, but no more. If something isn't directly related to your work, don't discuss it.

One more thing. Your audience can change depending on the venue. For example, the audience of a scientific article in a journal should be much more advanced than the typical audience of an MQP, IQP, or thesis. You don't need to explain as much, since the reader of a journal article should have more background in the content area.

7. Writing a scientific document is like writing a story. A scientific paper is much more than a "data dump" where you simply report the data you collected. It is really like to a story. Stories have a beginning, middle, and end. They also have conflict and resolution. A good scientific paper has important questions and answers to those questions.

You should first start out with the motivation and background to your work. Why did you do the research and why should I care? What is the problem you're trying to solve? All this information should go in the Introduction/Background. Also, remember tip #1 above: explain as necessary. Next you describe your approach to solving the problem (Methodology). Finally the climax comes: how did you solve the problem and what answers did you get? This is your Results/Discussion section. This is where you put all your evidence, arguments, and data for your answers. Always keep in mind that you need to support your arguments with evidence, either your own data (which you collected) or other's data (from citing other's papers). The ending of the story is your Conclusions section. What were the questions, answers, and outcomes of your work? Put them here where you summarize your work.

8. Include plenty of references. Research is not performed in a vacuum, but builds upon previous work and efforts. You will not be the first person to work in an area or on a research topic, but will build off other's work and efforts. References (or citations) are a way to link your research to other's research and show the reader you know what you're talking about.

Make sure to include plenty of references that support your ideas. If something is not obvious or common knowledge, it should have a reference. For instance, you don't need a citation for stating "Fuel cells produce electricity." since this is common knowledge. However if you state "Ir-Sn-Pt alloys are the best catalysts for ethanol oxidation." then you should have a reference (or multiple references) supporting this claim.

Also, using a program like Endnote will make your life much easier.

9. Pay attention to grammar and style. Scientific writing is different than "normal" writing. It tends to be more formal with specialized language. It is both precise and simple.

Use past tense. Use first-person. Use complete sentences. Avoid informal writing.