I decided I wouldn’t write a post about explaining your own research because doing that should be straightforward. There’s no one who knows your research better then you. And since research can be on absolutely anything, it’s not possible to give any real help with how to write this section. The only two pieces of advice I have are: use pictures if possible and the KISS principle (keep it simple, stupid).
So I’m moving on to tips on writing your methodology and/or analysis of results sections. Depending on your research, and how you go about validating/testing your research, you may need one or two sections for this. If you have to run a complicated study (for example a user study) that will take more than a paragraph or two to explain, you should probably separate the explanation into its own section. Otherwise, if you can explain fully what you did in less than 2 paragraphs, then you can probably just put in at the start of the analysis section.
Either way, before you present any of your results, you must explain clearly to your readers how you got them. Did you decide to do a user study? Did you run them against or compare them to previous solutions? Did you run a series of samples yourself and are presenting the differences? The reader should be able to determine what type of results your going to present based on how you decided to test your research. The description of what you did should be detailed enough that a reader could attempt to replicate your results.
When presenting actual results remember that pictures are key. If it’s possible to clearly and concisely show the results in a table, graph or diagram, then do so. When it comes to paper, the “a picture’s worth a 1000 words” is very true. Anyway you can make the information easier to understand and analyze the better. While you do want to write your own description of the results, it’s good if the reader can look at the values themselves and compare them and confirm what you’re saying.
Even if you include a table of raw data or a graph of results, you still need to write a section detailing the conclusions that can be drawn. When you’re writing this you want to focus on two things, a) highlighting the obvious big results (but don’t spend a lot of text on these) and b) pointing out interesting trends or secondary results that readers may miss. Because you are the one most involved in your research, it’s likely that you’ll be able to look at your results and go “oh cool, even though this isn’t statistically significant, the trend that these results are presenting provide weight to the theory …” These are the type of details that many readers of your paper may miss, and so you want to detail them as well.
For example, I did a user study comparing two user interfaces. We were hoping that our new one was easier to use (it was). All users used both interfaces, but some used the new one first and others used the old one. One result was that those who used the hard one first, were able to do more on the easier than those who used the easier one first. But this is somewhat expected. On the other hand, we also discovered that those who used the easier one first, were able to do more on the hard one then those who used the hard one first. This suggests that users were able to transfer the skills they learned on one interface to the other, suggesting learning is taking place.
The goal of your analysis of results section is to point out what interesting information your results contain and why. So if your results didn’t go the way you expect, see if you can glean any details from the results as to why. Is there something wrong with your experimental design? Did you miss an important factor? Is there something else you can learn from your data that you didn’t expect? Often the most interesting details are not those that answer your primary research question, but the extras you get for free.