PhD Defence Essentials

PhD defences are interesting beasts that are difficult to tame, yet they are essential aspects of academia. Building a healthier relationship by understanding its role in the academic courses is often a missing piece into the academic puzzle students are trying to piece together.

Introduction

Throughout my studies, I gathered a document including a series of questions that were asked to students defending their thesis (mostly in biology). After a couple of defences, I was able to start summarizing the information in a series of thematic questions. What I learnt so far is that a defence (including the proposal, the presentation, and the oral examination [i.e., the questions asked]) is:

1. Your statement on a particular topic, akin to a court case. You have to "prove" your point with evidence.
2. Humble in essence: nobody is perfect and no research is perfect. Recognizing the limits of your work as well as its importance and position in the literature (in your field, but also to a larger audience) is crucial. Willingness and openness to learn more is key.
3. Should cover all aspects of wellbeing and success for a future job. i.e., demonstrate
   • leadership ability (finances, ethical & professional conduct, project management),
   • working with others (equity, diversity & inclusion),
   • communicate ideas (public speaking, teaching, writing),
   • solve problems (analytical & critical reasoning, entrepreneurship, innovation & creativity),
   • expand expertise (subject knowledge, tools & technology),
   • be well (healthy living, resilience, self-knowledge),
   • plan career (career knowledge, apply & interviewing, networking & job search).
4. Basically, a thesis is more about "responsibilisation" or making one accountable rather than being defensive about the thesis. You are in a way "intellectually responsible" (or intellectually liable) for the product of the graduate studies you are doing. Thus, the questions asked in the "defence".

The list might seem long, but in all the defences I've seen, there has been a question on almost every single aspect here. In the eye of a researcher, these are all qualities that could lead to success. Source: McGill's myPath IDP workbook.

Preparation for the evaluation

• Make a document where you write notes on the core concepts you tackle in the thesis, you can add notes to this document while studying for the defence, but keep it concise, try explaining some concepts by drawing them, anticipate questions and answer them (get inspiration with the list below),
• Read your thesis, understand the core theory, hypothesis you are testing or answering, topics, methods, results, conclusions (perhaps search for other explanations), what are your contributions to the field, what are the limitations or weaknesses in your research (can this influence the results and conclusions of your studies?),
• Know your audience (evaluation committee): Check the websites of each member of the evaluation committee to know more about them, read a couple of abstract and papers from your evaluation committee to see their interests,
• Practice your presentation: (multiple times), make it engaging and clean, do practice talks with your lab, friends, family, etc.,
• Mock defence: Seek feedback by planning a mock defence with your peers, ask for feedback especially on the science that you are presenting (often, comments will be presented about the look of the presentation, but ask specifically for comments that will strengthen or clarify the arguments in the thesis),
• Additional visual aid: In your presentation, you can generate new figures to explain the concepts you are talking about, use figures for other papers (and cite them). Place all figures and supplementary information in a Slideshow (PowerPoint), at the end of the presentation you're going to give, to reference some ideas real quick,
• If you’ve published papers, read the reviews,
• Review the literature: Read major (hopefully recent) reviews about the global subject you study, try to find out how your research fit into the current literature,
• You can also watch some videos in your field on YouTube,
• Prepare physically take care of your mental health: exercise, have good sleep habits and eat balanced meals. Books can help you manage your stress. Presenting your research can be nerve-wracking. However, there are ways that can let you return to a calm and confident state. Please, learn these tricks as they will be helpful in other circumstances.
• During the defence, listen carefully to the questions and take your time. The defence is a dialogue among experts. You are an expert at this stage. So take the time to reflect on what you are going to say. If a committee member comment on something, you can still respond to a comment. When answering questions, aim to be concise and thoughtful. You are evaluated to become a peer at the same level as the other committee members. This means that you should respect some professional norms. Arrive ahead of time and test the projector, clickers, etc.
• Whatever the outcome, take a moment to look back on your accomplishments! You are at an amazing step in your career!

Attitude towards the questions

Defences are about having a discussion about your work and the field. Usually, the discussion is fuelled by challenging questions. We need to consider the types of questions and our attitude towards the questions asked. Think about the points below, especially for one of your future conferences.

1. Not all questions are clear. You can ask the speaker to reformulate.
2. Questions are sometimes asked because the speaker wants a specific answer or wants you to give nuance or a short summary of the subject to put in perspective. This is hard if they try to hide information that might ring a bell to you. Be vigilant of such "theoretical" questions.
3. In other instances, questions are legitimately bearing no answer. It is OK if you don't know everything. It is still allowed to show interest and try to imagine how it could be incorporated into your studies.
4. The questions (for a defence) should not only be about clarification, but to discern the limits of your knowledge.
5. Make sure that you understand the MEANING of the WORDS used in the questions. e.g., it's not because you haven't FOUND something that it doesn't EXIST! But you can say something like "With the CURRENT experimental design, I didn't find anything".

Thematic questions:

1. How to?

1. How to test for something (statistically, biologically relevant, theory, hypothesis, etc.)
   • What biological process (or biological mechanism) could explain the hypotheses? What could be an alternative explanation? Could you hypothesize a possible alternative and their mechanisms?
   • What unstated assumptions do you make?
2. What would be an experimental design to test [...hypothesis, biological processes, theory, etc. ...], given you have the resources?

2. What if?

1. Let's say I'm a grad student coming to your lab and [...ask you for something in particular about your subject...]
2. If [... definition of a term or interpretation of a model output ...] means X, does that imply Y (tentative explanation by the speaker to see how you would interpret it). (BE CAREFUL about these types of questions. Don't assume that the meaning of X is true, you can challenge the validity of the question).
3. If you had more money, time and other resources: what would you do?
   • What type of protocol would allow you to have even more robust samples and/or experimental design (i.e., do a common garden experiment, increase the sample size to do the more robust test X, etc.)?
4. How would your results change if process X had an impact on you study subjects? E.g., if climate change [or other processes] is affecting your populations [or organism], how would that change your results? (expectation?)
5. Tell me why I'm wrong on X. (requires explanations and nuances)
6. If you were to go back (wind the clock back!), and if your focus was different, what would you have done.
   • Would you have made the same choices in terms of site selection or species you've looked at?
7. Your focus is about "natural" phenomena. What about anthropic phenomena? Could they influence your results?
8. If the knowledge on assumption X changes (because a study contradicts the assumption you make), would the conclusion be changed?

3. Define and Explain?

1. What do you mean by sentence X? (e.g., I have an exceptional system; why is it exceptional?)

2. Define a concept (trait, character, evolution, population, phenotypic integration, countergradient variation, cogradient variation, local adaptation vs population structure…)

3. Why (or explain) you decided the particular values in your protocol?

   1. How was that informed (what kind of informed guessed you took?)
   2. How robust is your protocol? What would a slight modification of your protocol would have on your results? What other decisions could have been made to make the protocol even more robust (or get a more robust answer)?
   3. Why did you select these sites, species, groups, populations, etc.? Is this decision based on ecology, on a scientific decision, for a specific scientific inquiry? Or just feasibility?
   4. Why did you measure X and not Y? How is X possibly affecting Y or the other way around? Have you thought about the correlation between X and Y? Is X or Y symmetric?
      1. Is the sensory perception of your organism captured in your analysis (i.e., if your animal sees colour in UV and you haven't measured in UV spectrum, why you decided to drop that type of measurement? How could that impact?)
   5. Have you thought that just the sheer number of people working on organism X, the researchers could influence the results of your study on this organism?

4. Explain the result of an alternative test

   1. What would be the advantages of using method X instead of method Y (review vs meta-analysis)? What are the limits of method X?

5. Explain the caveats of a technique

   1. What are the (hidden) assumptions?
   2. What are the limitations? Can you think of limitations in your study that could change the conclusions or interpretations?
      1. e.g., limited number of REPLICATES, lab vs natural conditions, contamination, sample size, resources, etc. May affect the power of your analysis. Can you comment on that?
      2. Is your experiment representing realistic or a natural response or something in the environment?
         1. Are realistic are the initial conditions?
      3. Important to reflect on limitations: they could inform what you do in the future (because you'd like to explore these limits to guide your study!)
   3. How to avoid the limitations, caveats?

6. Explain the statistical tools

   1. Explain the models, meaning of results, or define the R squared, sample size, power analysis, effect size, variance explained, accounting for other variables, controlling for other data,

   2. Multivariate vs non-multivariate analysis, why select one more than the other?

   3. What do I learn (meaning) from this value or this analysis? Explain what it means to have an effect size of X in this a population?

   4. Statistical sampling (population to sample)

      1. Is your sample representative of the population? How do you know that?
      2. How balanced were your groups? Can this affect your analysis? What would happen if the proportions were different between the treatments?
      3. Are your results independent from each other?
      4. Have you assessed the variability within treatment? How is variability affecting your results?
      5. Do you have controls (positive or negative) for your experiments?

   5. How did you select the variables that you selected? Why not others?

      1. How to filter data to reduce the noise.
      2. Have you controlled (inclusion of covariates) for X (centroid size, size, sex, population size, distance, phylogeny, etc.)?
      3. Why would you size correct ? (Allometry)
      4. How do you account for the error? How confident are you in your estimates and your results?
      5. Do you think variable X could be related to ANOTHER unmeasured variable Z?
      6. Could you give me an alternative trait or variable I could measure to study the phenomenon you are studying if the variable you propose would not be possible for me to sample?
      7. What could be a real unmeasured confounding factor in your analysis? Would you advise someone to take that into consideration? Why haven't you used that?

   6. What are the predictions of model or theory X (pace of life syndrome, evolution by natural selection, neutral theory, etc.)

   7. If you've developed a NEW test or new method on the field or in the lab, can you justify your test or explain how it performs?

      1. Would you advise people to use it or not? How could it be improved? What are the assumptions and limitations?

   8. You ran a lot of models: correction of multiple testing?

   9. You account for variable X as unchanging through time. But is it the case and how would that variable change the outcome of your study?

   10. How many people collected the data?

       1. have you taken into account measurement error?
       2. Have you computed repeatability (even if 1 person did it, is that person repeatable?)

7. Biological explanations

   1. How confident are you in the validity of theory X in biology? Do you think it is widely supported?
   2. What mechanism could have caused X (increase in nucleotide diversity, would that lead to more resistance in the future?)
   3. The evolution, ecology, and development of your organisms:
      1. Could you explain a biological mechanism that would determine the underlying pattern that you observe?
      2. Consider the development, ecology, and evolution of your organism (difference between juveniles, adults, males, females, behaviour, local, time…)
      3. Can the results be interpreted within the ecology of your species? Does that make sense?
      4. Can you explain to me the biological origin of your organism? When it diverged, if it hybridizes, etc.
      5. How many generations have passed in your study? (Is there any evolutionary implications?)
         1. How much is additive genetic variance vs plasticity important in the study?
         2. from the analysis of X, how is this answering anything about the predictability of evolution?
   4. Explain the biological importance of what you found?
      1. How biologically meaningful is X? What is the effect size?
   5. Is your organism a good system to study what you wanted to uncover?
      1. Strength in using your organism, but are there problems (or limits) in generalizing for the tree of life?
   6. Multivariate nature of X (what other abiotic or biotic factors could be at play here that you haven’t measured or put into the analysis): what else would covary: oxygen, pH, temperature, turbidity, dissolved organic matter.

8. Reproducibility:

   1. How general are your results? How would they apply in other taxa?
   2. What are the spatio-temporal determinants of your populations? You've taken a snapshot of data, but how does that fit into the dynamics of the ecosystems?
      1. What is the link between the different spatial scales (individual, habitat, ecosystem, biome) and the ecological variables of the different environments?
   3. Temporal: What do you think you’d found if the study had been tracked for more time? Extend the study for more than X amount of time. To find if it is predictable or reproducibility. Do you think it is reproducible?
      1. You measure something that passes through time: Do you think that what you found is an equilibrium or that it is a transient state, assuming the conditions stay the same?
   4. Space: Is the result repeatable across space?

4. Who cares?

1. Why study this? (Beside having a job)
   1. what was the reasoning in doing chapter X?
   2. What is your elevator pitch?
      1. What is the elevator pitch of chapter 4! (if you only have 3 chapters.)
   3. What future avenues benefit based on what you’ve done? What are other things that can be studied based on what you’ve done (like a scientific research program)?
      1. False discovery rate: p-values, confidence on the data you have to continue and further the research (is it worth it to continue research on this subject?)
   4. Experiments relevant to the real world: What is the incentive of doing the research?
   5. If your study system is VERY much studied
      1. After all those years of study on organism X, why bother (and spending more money) to continue studying X?
      2. Do you think we start to see a big picture from all the literature that has been produced in your system?
      3. What is the priority or what should be studied on this system that would answer a theory or hypothesis? Basically, could you spearhead a lab that would have that aim?
2. What is the most important question in [... Insert your study subject or field of study, e.g., evolutionary biology ...]?
3. Why is [... Concept here ...] important? Is there extensive literature on this [... insert subject here ...]?
4. What do you respond if I say that what you study is not interesting?!
5. What new key insights have you brought into our field? Do you think, in your PhD, your question is a legitimate biological question?
6. What your study means for the:
   1. Public
      1. How is this relevant to society?
      2. What message would you bring to the public (elevator pitch) from your studies about your field of study/biology?
   2. Government
      1. How do you help to fund?
      2. How successful is the measure you propose?
      3. What information do you need ahead of time to fund those programs?
      4. If you have the chance to convince a politician of the importance of the work you do, what would you say to him?
   3. Conservation agencies:
      1. If you spoke with a conservation biologist, how would you explain your results so that they are applicable to base decisions on? What would be those decisions about?
7. Are the results applicable elsewhere?
   1. How is your work translatable to other contexts?
   2. What are the policy implications of your study?
   3. Inspire future research from your work, more power with another variable (what else should be done)

6. Select, remove, add Chapter

1. What is the chapter you like THE LEAST! and why?
   1. Not because you could do this differently: complications
2. What is the chapter you like THE MOST! and why?
3. What bad happened?
4. what would you have done differently?
5. If you had another chapter to your thesis, what would it be?
6. Are your chapters constitutive?
7. If you had a chapter XXX, what would it be?!
   1. Unlimited funds and human power:
   2. What would it be? What would be the grand question?

7. Why are you here?

1. What got you in science, why are you here?
2. What have you learnt the most in your studies? What advice would you give to improve all of us?
3. What is most needed in science to you could bring?
4. What did you want to do at first when you started your PhD?

8. What next, what’s your direction in the future?

1. What are your next priorities to weave together these wonderful stories?
2. Postdoc? What are you interested in? Phenotypes to test? More replication?
3. Where are you going next?
4. What the main anchor (to what group you affiliate) that you're taking from this.
   1. How do you define yourself? Evolutionary biologists? Guppy specialist? Trinidad expert?
5. Postdoc: what are you doing for that?

9. Miscellaneous

1. What advice would you give to yourself when you started the study?
2. What advice would you give to incoming students?
3. What advice would you give to someone if they wanted to continue your subject?
4. What are you most proud of?
5. If you had your organism in front of you and it could talk your language, what would you ask it?
6. Since you took some years to complete your studies, there might have been new publications since you’ve worked on your stuff. What update in the field can you make since you made your studies?

Conclusion

This list is not comprehensive, but might give you ideas on what you want to work on, or at least find the blind spots and explore these. By revising these guidelines and thorough preparation, I hope you have ideas on how to navigate your PhD defence successfully.

Acknowledgment

Gratitude to all who helped me in my studies and inspired me to learn more. You are welcome to suggest edits to the list of questions.