Ten steps to producing a well-written evidence-based health care dissertation

Step 1: Planning: “Don’t rush in hoping that something will emerge out of the fog”

Take time to plan what you want to write is a key lesson Carl has learnt. This is particularly important in the introduction, background and discussion sections of your thesis as they provide an opportunity to show off your writing skills.

Kamal explained that a well-written introduction and background gets the dissertation off to a great start and is your opportunity to demonstrate that your research question is important; that you have taken the time to read and understand the existing literature in the field; and have a clear vision of what your project adds. Make the case for why your topic is an important healthcare issue and why the methods you are using are appropriate for answering your research question.

Carl recommended using Powerpoint slides or post-it sticky notes to jot down the key points you want to make. You can then play with the order until you hit on a sequence that has a coherent, logical message.

 Step 2: Keep it simple: “Short sentences are far more effective”

Annette highlighted the pitfall of thinking that scientific writing means long complicated sentences. Instead, she recommended using simple, jargon-free language to express your ideas succinctly and clearly. Although technical language is important and essential, keep in mind that your dissertation needs to be readable by a broader scientific audience who are not specialists in that particular subject.

Excessively long sentences with multiple clauses make it harder to follow what you are trying to say. Try reading what you have written aloud: If you find yourself running out of breath before the end of a sentence, it’s too long!

 Step 3: Signpost the reader: “Make it easy to read and follow”

Carl and Annette both emphasized the importance of a clearly structured thesis that uses sections. It demonstrates to the examiner that you are organised and systematic – crucial qualities in a researcher! For example, if your project is comprised of a survey and focus groups, deal with them in the same order in both the methods and results sections.

Kamal indicated that students are expected to refer to reporting guidelines (see the Equator Website) for their chosen study design throughout the process of writing up their thesis. These will help inform the subheadings you use within each section to orientate your reader.

Step 4: Writing your methods and results

Each section of the thesis should follow on from the previous one.  Carl explained how marks can be lost when the reported methods don’t follow on from the research question previously defined.  Similarly be sure that the results you present arise from the methods you have described.  If you find yourself presenting additional methodological details in the results sections to help explain the findings, you may need to revise your methods section.  As Carl succinctly put it, “results are results, methods are methods!”

These sections can be time-consuming to write due to the amount of sometimes very technical information that has to be conveyed. Don’t forget your audience – provide them with the information they need to interpret your results in the context of their clinical practice. Think about which summary statistics are most meaningful to them.

Annette described that a common error is not using tables and figures appropriately to report results, instead describing them – very wordily – in the text. When designing tables aim to produce tables that are straightforward to understand and that don’t require additional explanation.

Step 5: Be reflective: “Present an unbiased view at the start and present an unbiased view at the end”

Kamal urged students to be reflexive and self-critical throughout their dissertation. The background section of your thesis should provide a balanced overview of the existing literature, not one biased by your own experiences or prejudices.

The discussion section offers you the opportunity to acknowledge the limitations in your work. There are things that it may not have been feasible to achieve in the course of an MSc project. Carl stressed the importance of “bomb-proofing” your work by acknowledging the limitations. Don’t think that by not mentioning them the examiners won’t notice! Instead be open and transparent, key characteristics of evidence-based health care.

Step 6: Editing and redrafting:  “I’m sorry I wrote a long letter, I didn’t have time to shorten it”

Carl quoted Mark Twain to illustrate his point that it’s a mistake to assume that a first draft will ever be sufficient. He advised putting your work aside, perhaps overnight, and come back to it with fresh eyes in the morning. It can be hard to edit what you have written – especially given the time invested– but the clarity and relevance of every sentence matters, particularly in the context of a word limited dissertation.

Step 7: Proof-Reading

Whilst your thesis must be your own work and the intellectual content your own, Annette suggested getting a draft of your thesis proof-read, perhaps by a friend or relative. By having someone who does not have specific technical knowledge read it, it will help you assess if you’ve written it in an accessible way. Furthermore, a dissertation with spelling and grammar errors can often detract from your work.

Kamal explained that supervisors are not expected to dot all the i’s and cross all the t’s but they can provide a steer if further attention is needed with respect to the proof-reading.

Step 8: Writing well takes time: “It’s such a cliché but it’s true”

No dissertation guide would be complete without mentioning time management! Much of the advice given by Carl, Annette and Kamal requires time. Kamal described how important it is to recognise that an MSc thesis is a substantial piece of work. He recommended setting yourself internal deadlines that you agree with your supervisor. Don’t leave everything to the last minute, especially if you want to receive feedback from your supervisor and/or proof-reader.

Step 9: Make it look good!

There is nothing more inviting to an examiner than a well set-out, consistently formatted thesis. Simple issues like ensuring that your line spacing is uniform and your tables are of a publishable standard will help gain those last few marks that could make all the difference.

Both Kamal and Annette highlighted that students should make the most of the training opportunities available to them. For example Oxford University IT Services offer a range of in-person and online courses and resources to help you produce your thesis.  You can explore what’s on offer here.

Step 10: A few final tips

• Avoid using informal language; for example, use “did not” instead of “didn’t”.
• Be consistent, for example in your use of decimal places.
• Use the active voice where possible, rather than the passive voice, as this is often more wordy.
• Don’t forget there are examples of well-written dissertations on weblearn.
• Your dissertation handbook is a good place to look for general advice about your thesis and formatting and referencing guidance.
• You can read more about the University’s proof-reading policy here: http://www.admin.ox.ac.uk/edc/policiesandguidance/policyonproofreaders/

Many thanks to:

Professor Carl Heneghan, Director of Programmes in Evidence –Based Health Care

Dr. Annette Plüddemann, Course Director: MSc in Evidence-Based Health Care

Dr. Kamal Mahtani, Course Director: MSc in EBHC Systematic Reviews

Image courtesy of A Mentor.