On Twitter this week two people asked me for advice for starting the discussion chapter of their thesis / dissertation (I’m going to use the word thesis from now on because I am Australian). I didn’t feel up to answering in 140 characters or less, so I promised a post on it today.
If you are feeling anxious about the discussion section rest assured you are not alone. It’s an issue that comes up time and time again in my workshops. There’s no one answer that can help everyone because every project is original, so I thought I would offer a few thoughts on it by way of starting a conversation.
Evans, Gruba and Zobel, in their book “How to Write a Better Thesis”, describe the discussion chapter as the place where you:
“… critically examine your findings in the light of the previous state of the subject as outlined in the background, and make judgments as to what has been learnt in your work”
Essentially the discussion chapter tells your reader what your findings might mean, how valuable they are and why. I remember struggling with this section myself and, looking back, I believe there were two sources of anxiety.
The first is scholarly confidence. At the University of Melbourne we used to talk about how a good thesis has a ‘Ph Factor’. The Ph factor is somewhat elusive and hard to describe, but basically it means you have to make some knowledge claims. You need to have the confidence to say something is ‘true’ (at least, without getting too post modern about it, true within the confines of your thesis). This can feel risky because, if you have been approaching the thesis in the right spirit, you are likely to be experiencing Doubt.
The second source of anxiety is the need to think creatively. Most of the rest of the thesis asks us to think analytically; or, if you are in a practice based discipline, to make stuff; or perhaps, if you are an ethnographer, to observe the world in some way. Creative thinking involves your imagination, which means you have to switch gears mentally.
So the problem of the discussion chapter is a problem of creative thinking and confidence, but there are some stylistic conventions and knowledge issues that complicate the task. Every thesis needs to have discussion like elements, but they may do it in different ways.
In a conventional thesis, what we call the IMRAD type (introduction, methods, results, discussion and conclusion) the discussion chapter appears a discrete chapter. Before you worry about the discussion chapter too much, consider whether you need to treat the discussion as a separate section at all. You need to keep in mind that the IMRAD structure is best used to write up empirical research work (the type where you collect data of some kind).
In the past I have referred to the IMRAD formula as the ‘dead hand of the thesis genre’; a phrase I picked up from my colleague Dr Robyn Barnacle. It’s a dead hand because of the role it plays in the imagination of the research community throughout the world. The IMRAD formula is the most widely understood format because it is the type most widely described in the ‘how to’ genre and has a close and abiding relationship to the scientific method. Many students try to make their research fit into the IMRAD format, when it is not appropriate to do so.
I can be easy to feel ‘blocked’ if you are a non scientist trying to separate out the discussion from the rest of what you are writing. Remember there are many ways to skin the discussion cat. For example, an artist may discuss each project and what it means separately. An ethnographer might devote a chapter to each theory they have built from observation. Likewise a historian may break the thesis up into time periods and do critique and evaluation throughout the whole.
So I have diagnosed some of the problems, are there any easy solutions? Well, the best way to start in my view is just to write, but perhaps start to write without the specific purpose of the discussion chapter in mind. Write to try and work out what you think and then re-write it later.
You can use a couple of basic techniques to help you with this process:
- Try the old ‘compare and contrast’ technique. Draw up a table describing where your work is similar to others and where it differs. Use each of these points as a prompt to write a short paragraph on why.
- Use the “The big machine” trick as suggested by Howard Becker in his book ‘tricks of the trade’ (now only $3.99 on Kindle? Bargain!). Pretend your results are produced by a machine then describe the machine. How would the machine work? What would it look like? What parts would it need? What might make the machine break?
- Another useful suggestion from Howard Becker is the null hypothesis technique; write down why the results mean nothing. Sometimes forcing yourself to argue the reverse position can highlight the relationships or ideas worth exploring.
- Sometimes having an audience can help. Explain the results to a friend and record yourself, or use voice recognition software to tell your computer some of your preliminary thoughts. Many people find talking an easier way to get ideas out. Alternatively write them in an email to someone.
- Explain the limitations of the work: what is left out or yet to do? Sometimes, like the null hypothesis, talking about the limitations can help you better define the contribution your study has made.
I hope some of these suggestions help to get you started. Do you have any more? Are there ‘tricks’ you have used to help you get your creative juices flowing?
The Dead Hand of the Thesis Genre?
Ambivalence: can it help with your PhD?
You finish your dissertation with a conclusion and a discussion. These are two different sections or parts, but sometimes it’s difficult to separate the two. This article discusses these two parts and the information that each part must, at a minimum, contain.
What does a dissertation conclusion look like?
The purpose of a conclusion is to answer your research question. Begin, with repeating your research question. However, don’t simply reiterate the research question, but integrate an explanation of it into the rest of the section’s discussion.
Then give the conclusions that you draw based on the results of your research (use the key results that are most relevant for answering your research question).
Finally, answer the main question and explain how you have come to this conclusion of your dissertation. Don’t just list the question with the answer below it, but carefully explain it and incorporate it into the rest of the text. Provide the raw observations and don’t interpret.
Example 1: Provide raw observations
Raw observation: Ten respondents agreed with the proposition that elephants are wonderful animals and six disagreed. Thus, the majority was in agreement.
Interpretation: Ten respondents agreed with the proposition that elephants were wonderful animals and six disagreed. This shows that ten people were very impressed by the animal’s character and that six people hated elephants.
In terms of verb tense, in the conclusion you use the present (simple) tense when you present facts, and you use the past tense or present perfect when you refer to the research that you have done.
Example 2: Verb tense in the dissertation conclusion
Facts in the present (simple) tense: Company X currently has no clear vision and mission. The company also does not make (is not making) good use of social media to sell its products.
Your research in the past tense: This research examined whether Company X does have a clear vision and mission […] The results showed that Company X … […] The researcher also analyzed when …
Your research in the present perfect tense: This research has examined whether Company X does have a clear vision and mission […] The results have shown that Company X … […] The researcher has also analyzed when …
The length of the dissertation conclusion should be between 200 and 400 words. Pay attention:
- Did you use a hypothesis instead of a research question? Then indicate whether the hypothesis holds true.
- Do not provide any new information in the conclusion. Any new information belongs earlier in the dissertation.
- Do not give examples in your dissertation conclusion, because you should have already worked out your conclusions (with examples) in the rest of your research. For example, if you conclude that the financial crisis has negatively affected the revenues of banks, then you should not also mention that Bank XYZ has had, ‘for example’, 20% less revenue in 2009 than in 2007.
* Use of the first person ‘I’ is often incorrect and prohibited in many studies but is accepted in some studies. Find out how your study treats the first person before you use it. In any case, we recommend that you don’t use the first person anywhere but in the Preface. Do you want to know more about this subject?
What does the discussion look like?
In the discussion, you write more interpretatively and colorfully about the results. Whereas you kept it concise in the dissertation conclusion, you write more in-depth about the subject in the discussion section.
Here you evaluate the research: you may discuss your expectations of possible causes of and consequences of the results, possible limitations and suggestions for follow-up research. Write the discussion, just as the dissertation conclusion, primarily in the present tense.
Interpretation of results
Start your discussion with the validity of your research design. Then discuss the results and indicate whether they meet your expectations. In this section, you will give explanations for meeting or not meeting these expectations.
These expectations can come from relevant literature, but they may also be based on your own common sense. In any case, describe how your results fit with the framework that you have drawn in the first chapter (introduction, motivation, theoretical framework, and research questions or hypotheses).
Also show how the findings provide new or different insights into what was already known. Elaborate on all possibilities: what exactly have you now shown?
Limitations of your research
Present the limitations of your research in a new paragraph within the discussion. Describe which observations you can make based on the research results. These remarks may be of an advisory nature.
If there are some side notes that can be made to the research or you were hindered by certain limitations, these issues can explain of the results you obtained. Name these, but also explain how these factors can be improved in future research.
Pay attention that you don’t run down your entire research project; you’re not supposed to provide a list of all your errors. These should have been carefully considered before you started your research.
Recommendations for further research
The discussion ends with a paragraph of suggestions for possible follow-up research. How can other researchers build upon your research? Do not write something along the lines of ‘there is still much research needed’.
You’re not supposed to give a list of what others should look into for you in order to complete your research, but rather give suggestions for a separate, follow-up study in response to your research.