PhD Qualifying Exams, or Purgatory for Graduate Students

For many PhD students, summer is the time of year when classes slow down and research picks up. When campuses are quieter and happy hour is earlier. Summer is an idyllic time for most PhD students. Except for those who have to take the dreaded qualifying exam. At many institutions, qualifying exams happen during the summer. They take many forms – some students must pass true qualifying exams, where they show mastery of the Master’s level curriculum. Others take comprehensive exams, where they show mastery of the PhD level curriculum. Some take both, typically separated by about a year. In some fields, qualifying exams may be written; in other fields these may be oral exams. Sometimes PhD students gain candidacy (and cheaper tuition) upon completion of their exams, in other instances there may be additional hoops to jump through before this can occur. Ugh.

Qualifying exams are a rite of passage. It is hard to provide rationale for why they are necessary, other than: I had to do it so you do, too. OK – that’s not entirely true. There are good reasons for students to show that they can synthesize the information from several classes and utilize that knowledge to solve problems which may be more complicated than ones they might find in a classroom. In Biostatistics, there is often an a theory exam and applied exam, the latter which allows a student to demonstrate that they have mastery of important collaborative skills. Regardless of the process or procedure by which a student must prove that they are “qualified” to continue in the PhD, it is a very stressful time for most students. And rightfully so. Their performance on their exam could have long-lasting implications for their career. Not to mention their psyche.


Psyche: William-Adolphe Bouguereau (1892)

To complete my Master’s degree in Biostatistics, I needed to pass an MS comprehensive exam. This was also the qualifying exam for the PhD students, so my classmates in the first year of their PhD program took this same exam as I did at the conclusion of my master’s degree. The exam went moderately well for me, with the exception of the last question. As I completed the problem, for which it was necessary for me to compute a probability, with almost no time left, I was dismayed to find that my solution was negative! I clearly had made a math error somewhere – of course I knew that a probability had to be in the set [0,1]. I was almost out of time, so just made a note “I know this can’t be right since my probability is negative, but I am out of time and cannot go back to figure out where my error is.” When I saw the results of my exam, although I passed, I was disappointed to find that the Professor who graded it disregarded my note, and essentially dismissed my solution, with a rather degrading comment. This was the first time in my academic career I felt downright stupid. It had an impact on my confidence in my future studies. All of a sudden, I considered myself to be no good at theory.

And then, my PhD. I did very well in my classes and prepared like hell for my qualifying exams. At my institution, we had a 2-day in-class theory exam, followed by a 1-day applied exam. I studied with classmates. I studied alone. I ate, slept and breathed the central limit theorem. I was ready. This exam was mine. Except it wasn’t. The day of the theory exam came and I found myself hiding in the bathroom, hyperventilating. And did you know that you could play Minesweeper on the HP-48G calculator?

calcminesweeperEventually, I picked myself up, went back to the exam, and did what I could, which wasn’t much. And so I failed the theory exam. No one had failed the theory exam in so many years, that none of the current students could remember the name of the last person to have failed. I was dejected. I was humiliated. I was sure that I needed to consider my non-scientific career choice (an event planner, by the way… which I may have to reconsider now that I’m planning a Bat Mitzvah celebration). Lucky for me, I had the opportunity to take the exam a second time. I just had to decide if I wanted to. Could my ego handle more failure? Seriously: what would I do if I didn’t pass? I took long walks with my friends (now my colleagues), and talked about what my alternatives were. My husband and I considered what options I had. In the end, I decided to take the exam again. And long story short, I passed. As my advisor told me, I didn’t “do great”, but I did “well enough”. Well HOLY SHIT! As far as I was concerned, well enough was good enough to pass, and that was good enough for me to continue in the PhD. Woohoo! But the damage was done.

I continued in my PhD, and successfully defended my dissertation some time later. After that, I started a job as an Assistant Professor, and the rest is history. And like all history, we tend to forget the parts that aren’t as pleasant. However, I could never quite forget my experiences with my exams. For years, I would tell people that I wasn’t theoretical, that I’m an outstanding applied statistics, even while teaching one of the more theoretical doctoral level courses my Department offered. I descried my methodological work as applied, since it was motivated by real problems. The damage was done.

As time went on, I became the Director of Graduate Studies. Now it was MY responsibility to oversee the qualifying exams of others. This was excruciating for me. All of my anxiety would well up each time the exam came around. I would agonize over whether the questions were consistent with those we asked in previous years. I would lose sleep about whether we had sufficiently prepared our students to do what we were asking of them. When the exam was being administered, I would delegate a different member of the committee to proctor the exam and I would not step foot in the building. The students’ fear and anxiety were too real to me. The damage was done.

Like most negative experiences, though, I found a way to use my failures to help others. Unfortunately, we typically had a student or two each year who did not pass their qualifying exam. It was my responsibility to pass that news along to them. When I did so, I always shared my experience, and I believe it helped them to know that this negative did not define me. Even if I still struggled to believe this, I was able to convince them that it was true.

And now? Now I’m a Department Chair. It is clear that my qualifying exam experiences did not hinder my career. Do I think I will ever have a JASA paper? No. But that’s okay – there are plenty of other statisticians who can fill that role. I am pretty good at what I do, and there are many strong theoretical statisticians who are terrible collaborators. Does it still sting sometimes? Absolutely. People who get PhDs are over-achievers, and experiencing failure is not something we do often or handle well. However, learning from my failures, and using them to help others, is a skill that I’m so glad I have. I believe that my failures make me much better at my job today. I know that my failure helped me better understand my strengths and my weaknesses.

Could things have been different if my experience with my Master’s exam had been different? Maybe. Perhaps I would have been more confident in my theoretical abilities, which in turn could have led to me passing my PhD qualifying exam on the first try. But maybe not. I am a really good collaborative statistician, and I enjoy it. So it could be that I still would have found that path, even if I felt that I excelled in the theory.

In the end, I still hate qualifying exams and always will. I understand their purpose, I understand the benefit they provide to our students in the long-run (in a multitude of ways), and I know they’re not going away. Unless the Department Chair has the power to do that….



The Joint Statistical Meetings… Then & Now

In just over a week, I will be attending the 2016 Joint Statistical Meetings (JSM) in Chicago, the largest annual gathering of statisticians in North America. I can see my non-statistical friends cringing, but to me it is pure nerdvana! With over 6000 attendees and over 600 sessions, JSM can be overwhelming at best, and downright scary at worst. As I’ve started planning my time in Chicago, I realized that 20 years ago, I attended my very first JSM, also in Chicago, which caused me to stop and reflect on how much has changed in that time.


Twenty years ago, I was in the middle of my Master’s degree at the University of Iowa, a wide-eyed kid trying to figure out what my next steps would be. A group of my classmates and I piled into a car and made our way to Chicago from Iowa City. There were 4 of us, 3 girls and a boy, and we all piled into a single room, with my male classmate respectfully sleeping on the floor. Each night we would all review the program for the next day, amazed by all the different talks we could see, trying to figure out which matched our interests. It was during this meeting that I attended talks which sparked my interest in clinical trials, drove my desire to someday write a statistics in sports paper, and reinforced my professors’ repeated lessons that good communications skills are a vital part of being a statistician. I remember walking into rooms in which I was the only woman, amazed that this was the 90s and that could still happen! I encountered “famous” statisticians (yes, there are such people) and was in awe; I also encountered my first instance of sexual harassment, from which I learned a valuable lesson about humanity. I learned as much from those few days in Chicago as I had in a year in graduate school – different lessons, to be sure, but still a tremendous amount. Most of all, I learned how little I knew, how much more there was to learn, and promptly decided that I needed to stay in school and get my PhD.


Fast forward 20 years, and here I am again, heading to Chicago with 6000 of my closest friends. Now, as a Professor and Department Chair, my perspective is different. I still go to JSM with my eyes wide open, with plans to continue to learn. However, what I learn will be decidedly different. This year at JSM, I will attend a chair workshop, during which I’ll have the chance to learn how to do my job better from others who have been doing it longer. I will meet with colleagues from the National Math Alliance, an amazing organization with a goal of increasing the number of underrepresented students in PhD programs in the math sciences and be reminded of how important it is to have diversity in our profession. I will have the chance to catch up with former classmates and colleagues, during which time we will laugh at the people we once were, and talk about the people we hope to be. And I’ll get to listen to people talk about statistics for 3 days. Three wonderful days during which I will remember what inspired me to be who I am today, and reinvigorate my passion for statistics. I will be inspired to become a better statistician. Because, no matter what my job title, no matter what my day-to-day responsibilities, deep down, I’m still just a nerdy girl who likes statistics and who thinks that data are cool. And maybe I’ll get a good idea for that sports paper I still haven’t written.


Twenty years is a long time, and I’ve come a long way in that time, but I still have a long way to go. The experiences I bring home with me from JSM will help set me on the path for my next 20 years. And who knows, maybe in 2036 I’ll be celebrating JSM back in Chicago!

Why I love study section (and why you should too)

I recently returned from serving on study section. My academic friends probably read that and groaned, but for me, study section is one of the most scientifically interesting things I do. For my non-academic friends, let me give you a little context. Study section is a group of scientists brought together to review scientific grants, and give input to the funding agency as to which are of high scientific quality and should be considered for funding, and which should not. Each member of the panel receives several grants to read ahead of time, provides critiques and scores, and then everyone comes together for a couple of days to discuss each grant proposal and its merits and weaknesses. At the end of the day the researchers who submitted the grants will receive scores comments from all of the reviewers. The agency to which the grant was submitted then determines which researchers will receive the money they requested. The NIH has a very comprehensive description of the process here, and if you’re not into reading information prepared by the government, you can listen to someone else read you government information on a very dry YouTube video.

The person who organizes and runs the study section is called the Scientific Review Officer – the SRO. Every study section is run differently, and this is highly dependent on the SRO. I have served on several different review panels for different agencies (e.g. NIH, NASA, American Heart Association) and truly believe that the SRO can totally make or break the experience.

The study section I regularly serve on reviews grant submissions that are concerned with clinical trials evaluating therapies for neurologic diseases and disorders. We are often evaluating grants that are requesting very large amounts of money that could have a big impact on patient populations. Our SRO works extremely hard to put together a multidisciplinary group of reviewers who take this responsibility very seriously. These are some of the most intelligent people I have ever met. While some study sections only have 3 reviewers who read each grant, our study section routinely has 5-9 – this allows each person who reviews the grant to focus on their area of expertise. We have extremely thorough discussions about each grant, and we all hold each other accountable.


Our study section is very diverse – we have expertise in neurology, neurosurgery, pharmacology, biostatistics, physiology, and several other areas that come and go based on the science we are evaluating. We come from all over the country, each have different research backgrounds, and the group includes Deans and Department Chairs, researchers from industry, as well as Assistant Professors who may be reviewing grants for their first time. However, when we discuss the science, we all respect each other for their expertise, regardless of our positions in our real lives. We may not always agree, but we don’t view our discourse as a problem, rather as a learning experience. It is truly extraordinary, amazingly intense, and unbelievably exhausting. We spend very long days doing something for which we get very little reward, except for the satisfaction in knowing that we are helping to advance science (and a small honorarium which probably comes to pennies on the hour given the time we spend).

It is a lot of hard work. But I always look forward to it. My fellow members of the study section are also my friends. We share our stories with each other. We greet each other with our photos of our children out front, and with our egos behind us. At the end of the day, we share meals and drinks and laughs and try to unwind after some very tough discussions. We all go home having learned a tremendous amount from each other, and better scientists for it.

There is a lot of discussion in the statistical community about serving on content-area study sections. It is a lot of hard work and not all experiences are as positive as mine. But it is extremely important that we continue to bring our expertise to the table, and continue to build relationships with other researchers. In fact, it is our responsibility as scientists to ensure that all research is subject to statistical rigor that ensure good study design and results.

I am fortunate that I regularly serve on a study section that respects all opinions, even those of the statisticians. I am lucky that I serve on a study section that provides me with so much scientific satisfaction. And I am lucky to have served with so many amazing, brilliant people, from whom I learn to be a better scientist.