How to successfully defend your PhD thesis

A Ph.D. defense is one of those things that one can never feel prepared enough for because usually, it is the first time one is doing it! 
I go over mine here, having successfully defended my doctoral thesis in particle astrophysics a little over 48 hours ago.

Title slide of my defense talk. It says Studies in particle astrophysics with the ANITA experiment. Shows my name Oindree Banerjee, school name Ohio State University, the date July 12 2018

A defense is 2 hours long. You give a talk, answer questions, work stuff out on the board, and survive.

Usually, there is a public part of one hour where you could invite friends and family. 

I figured: it's an exam - it's go-time. 

I am "defending" means I must be, at least, somewhat under attack. 

If there is going to be war, I am ready and I didn't want distractions, so I told my friends and family not to come. 

It was me, and the 5 committee members. 

They started by closing the door and saying alright, we can torture you for two hours!

So, you know, it's up to you. 

Do you want no distractions but be alone with them for two hours or do you want some supporters empathizing with you for the first hour? 

I made a choice there and I am fine with it. 

The talk

As part of defending you write a talk that would leave plenty of time for questions. 

Using something like PowerPoint. 

My talk was 63 slides including the title slide (above), introduction slides, 6 backup slides, and 4 "divider" slides. 

Quick orientation

After the title slide, I included a "quick orientation" slide. 

I have done this in job talks as well and it helped. 

It tells people roughly what to expect.

quick orientation slide from defense talk introducing the experiment I will talk, detection technique used, who am I and names of my committee members


Then I do some motivation and introduction slides - I had total 7 of these. 

Especially making sure to motivate anything that is important to the research I will go over later. 

Below is one such slide. It shows where my experiment is on the energy scale as compared to other experiments/collaborative efforts.

All of the other experiments are either particle or particle astrophysics experiments. 

This slide serves multiple purposes. It is good to acknowledge other experiments and demonstrate knowledge of the field while pointing out why your experiment is unique. 

this is an example introduction slide from my defense talk showing where my experiment is in the energy scale as compared to other experiments

As part of my introduction slides, I explained the basic principle underlying our main detection method. 

Most questions 

This is the slide that I got the most questions on. 

This is the introduction slide I got the most questions on during my defense exam. It shows the basic principle behind the detection method used in my experiment

I was grilled on all parts of it to exhaustion and definitely feel that I could have kept things more straight here. 

I got questions on almost every introduction slide and they involved board work. 

I knew the committee was not trying to attack me but it was still kind of rough.

As a senior grad student, I had been spending all my time with research, not the basics and coursework. 

I had studied for the defense - I went back to undergrad and grad E&M and particle physics books, re-read class notes, etc. 

All that helped. 

But they still found spots that I was weaker on that day. 

Always learning!

The key was to get back on track with the talk once each grilling session was done. 

My brain was tired. It didn't help that I was up till 2 am the previous night. 

Then again, being tired is nothing unusual for a grad student and we can still put up a fight, right? Right. 

Bounce back and pounce again. Like nothing happened. 

This was challenging but had to happen in order for me to get through the presentation and show the stuff where I am the boss --> my research.

The experiment

I introduced the experiment that I knew and loved. 

Picture of the ANITA payload in Antarctica with myself and Linda standing in front. I showed this during my defense exam as part of introducing my experiment

Here I am with a postdoc colleague, Linda, in front of the ANITA-4 payload in Antarctica. 

I had 6 slides, including this one, on the experiment and how it worked. 

My stuff

The rest of the talk was my research, stuff I had done, and owned, and published, or not yet published - whatever, it was my stuff.

I used a divider slide to announce the start of a new section of research. 

I presented 4 main projects in my talk and put a divider slide before each one. 

In the divider slide, I put 2 pieces of information with the goal of providing a very quick summary. 

A title or brief description of the project I was about to present. 

And a hyperlink to the publication associated with the project if there was one out already. 

I put the web address to the published article in the hyperlink and edited the text that would show up to say the journal's name and volume number. 

Below is my divider slide for the start of the first main project I presented.

divider slide in defense talk
This project involved doing the collaboration's main analysis. 

This is the sort of project/paper that pays the bills.

So it's important, so I put it first. 

Another divider slide is below. 

divider slide in defense talk

This project is on the instrument and fortunately, what I worked on for the instrument greatly improved our livetime so I could talk about that.

I loved this project. It was mission critical and therefore, very exciting and rewarding. 

In this divider, I also added a hyperlink to a "trailer video" I have made on the project. The video is pretty entertaining so I thought I would include it in case there was a chance to show it.

The other 2 projects I presented did not yet have publications - they are the more recent projects I have worked on. 

divider slide in defense talk - blog how to phd

I only showed a couple of slides on this project. 

This was a short project in terms of the scope and mainly involved simulations. 

I enjoyed it and it was sort of optional - which is great for learning!

divider slide in defense talk - blog how to phd
This is the last project I worked on and the last project I covered in my talk. 

I was heavily invested in this project and driving it from the start, but got to work on it mainly at the end of my time. 

A slide from this section is below. 

defense talk example slide showing search design for ultra high energy neutrinos from Gamma Ray Bursts

This slide has happy memories from when I was in the UK doing some of the groundwork for this project. 

Like any normal physicist, the stuff that I dread getting grilled on is statistics - and, of course, that came up too. 

I got through my slides, said my conclusions, and answered more questions. 

The last question I got had to do with a rainbow and a puddle and Brewster's angle and I didn't get it. 

At the end

I went out of the room feeling terrible about the rainbow question, thinking: does this mean I fail. 

It was a few minutes before I got my result but felt like forever.

I sat in a chair down the hall. Such a strange and vulnerable feeling to be done as I could no longer do anything about it. 

I didn't want to see anyone until I heard the result. 

Students in the group kept passing by me as they - the kind souls - were setting up a celebration in the analysis room. 

I told them to go away. 

Then the committee called me in again. 

I always envisioned myself collapsing at this moment but ended up acting totally normal as they shook my hand one by one saying things like "congratulations", "great job", and "nice job."


  1. Hi....just came across your Blog...I am from Kolkata too...great job...very inspiring

    1. Thank you, Satyaki, great to see you here!
      Thanks for reading and hope to stay in touch :)


How to finish your PhD and graduate in a timely manner

So you started a Ph.D. Great! You are doing lots of research, or maybe you are burned out. Either is understandable. Maybe your a...