Reflections on deciding to leave for grad school

Roughly two years ago I decided to leave my job at IQS Research in Louisville and move to the DC area to attend full-time grad school at the Joint Program for Survey Methodology (JPSM) at the University of Maryland. This decision was hugely significant in my life and provided a learning experience in itself about making decisions. I’m taking the time here to reflect on it.

Why I Decided to go to Grad School

Accepting the offer was an easy choice, but deciding to apply to a grad program in the DC area was a difficult decision. I was happy with my job as an analyst at IQS Research and finally feeling at home in Louisville. At IQS, I got to work on interesting problems with clear real-world impacts, and entirely with coworkers and managers that I liked and respected. However, I knew that a graduate degree would be important for my long-term career. Most importantly, I knew that it would help me spend more time doing the kinds of work I most enjoy: finding ways to improve survey statistics that inform social policy. On top of that, I wanted to “get credit” for the countless hours I’d spent learning statistics, data science, and programming in the years after graduating from college in 2014. Financially, I expected the long-term benefits to outweigh the short-term costs. Sure, I might have to take out some student loans or face an opportunity cost of switching from full-time to part-time work. But over the next decade or so, I figured a master’s degree would be useful to IQS or any other employer I might work for, and I’d eventually recoup the investment.

Preparing for Grad School

I started preparing for grad school in 2017, without necessarily knowing when and where I’d actually enroll. I knew that if I was going to succeed in a statistics program, I would need to hone my mathematical skills and take two specific classes: linear algebra and real analysis. So I started out by teaching myself how to write (and enjoy!) mathematical proofs using Daniel Velleman’s wonderful book “How to Prove It”, as well as reading Lara Alcock’s insightful book “How to Study as a Mathematics Major”. Hunched over these books and a college-ruled notebook, I spent dozens of hours over lunch breaks and at coffee shops learning how to prove contrapositives and inductive statements by completing all of the exercises in Velleman’s book. While this was no doubt work, I came to love it, unlike when I had studied proofs at Amherst in a discrete math class taught by one of Velleman’s colleagues in the department. Back then, the art of proof-writing seemed arcane and mind-boggling. But working through Velleman’s book, I came to love the elegance of it.

In the three years before JPSM, I took a series of classes at the University of Louisville as a post-baccalaureate student, learning the math I would need to survive and thrive in a statistics program. IQS was supportive of my continuing education. They let me flex my hours so that I could go across town to attend classes during the workday, and they helped make sure my workload adapted to the seasonal rhythms of midterms and finals.

Two-to-three days a week, I would Uber over to University of Louisville in the afternoon and either Uber back to work or catch the bus back home. The bus stop where I would usually wait is in the photo below, which was the cover of the 2015 album “Little Neon Limelight” by the Louisville band Houndmouth, whose music I frequently listened to while waiting there. Eventually I bought a car, primarily to make it easier to schlep back-and-forth across town for work and classes.

(Pictured above) The bus stop I would catch at the end of the day after work and classes at UofL, as seen in the album cover of the Louisville band “Houndmouth.” I loved listening to that 2015 album “Little Neon Limelight” as I sat at that bus stop.”

In all, I took four classes there: linear algebra; an introduction to graduate-level mathematical statistics; and a two-semester course sequence of real analysis. I can’t tell you how much I enjoyed that class on real analysis. The professor, Dr. Alica Miller, was an exceptional instructor with a fascinating background. Her parents were both mathematicians, a Serbian Muslim and a Chicago Jew, who instilled in her an infectious love of math and a gift for helping others appreciate the poetic elegance of a well-written proof. She peppered her class with silly little jokes and anecdotes. By far my favorite joke was when Dr. Miller pretended to sic a giant athletic student on a pretentious physics student, shouting “Jason, crush him! Squeeze Theorem in action!” I also had a great experience with the applied math PhD students who mostly populated that small class. They invited me to join them for study sessions in their grad lounge, and we all benefited enormously from writing out chalkboard arguments for each other and commiserating about the tough time we had in that course.

Deciding Where to Apply

After finishing the last of these classes in the spring of 2018, I decided to apply to master’s programs in statistics that coming fall. Initially, I planned to apply to the University of Kentucky and to Georgia Tech. Both of these programs might allow me to continue working at IQS while earning my degree, and the Kentucky program would be fully funded.

However, that summer I fortuitously discovered that there was a master’s program specifically focused on survey statistics. During Shabbat dinner at a friend’s house, I met Cybil, a manager at the nearby Census Bureau National Processing Center in Jeffersonville, who shared some truly impactful career advice as well as a delicious recipe for Swiss Chard. I’ve forgotten the recipe, but I do remember her recommendation that I apply to the Joint Program in Survey Methodology at the University of Maryland. I’d never heard of it before, but she (correctly) suggested that the program would be a better fit for me than a typical statistics department. And though I’d never heard of it, she told me that it was well-regarded in the federal statistical system and provided excellent training in survey statistics.

Immediately it was clear to me that JPSM would be an excellent fit, but it was unclear whether I would have any way to pay for the degree without taking on student loans. So I initially decided to apply to JPSM, Georgia Tech, and the University of Kentucky, figuring that eventually if admissions worked out favorably, I’d have to choose between a potentially expensive, full-time program in exactly what I wanted to study versus an easily affordable part-time program with plenty of coursework that would be less useful to me.

I did all of my admissions work in the fall and winter. I studied a few hours each week for the GRE’s math section, took the exam a couple times, asked for recommendation letters, and wrote application letters describing my interest in each program. Writing these letters was a valuable experience. In having to explain my goals to others, I came to really understand what I was hoping to get out of graduate school. It also helped me understand why I was so interested in statistics and gave me a chance to reflect on which kinds of work I find fulfilling and energizing. The experience reminded me of the following quote about writing that I first heard from Dr. “Biddy” Martin at Amherst:

“Writing is thinking. To write well is to think clearly. That’s why it’s so hard.”

David McCullough

Writing was what made me come to feel so strongly about choosing the full-time M.S. degree at JPSM over the other options. I realized that if I was going to devote years and thousands of dollars (in tuition and opportunity costs) to a statistics education, it should be for an education that trains me squarely in the kind of work that I want to do for the rest of my life. Namely, statistical work that helps government and public-interest organizations better understand the populations they serve. The other two programs that I could attend while working at IQS would be significantly less useful in helping me do this kind of work, and attending them while working full-time would require either a long, soul-crushing commute between Louisville and Lexington or doing 100% online classes. Ironically, I found the prospect of 100% online classes very unappealing, but ultimately that’s what happened for the majority of my time in grad school during the Covid-19 pandemic. But anyway, back to 2018…

In contemplating the options of Lexington vs. DC vs. 100% online classes, I couldn’t help but wonder if I would like living in the DC area more than Louisville. I’d be closer to my friends and family on the East Coast. The cost of living would no doubt be higher, but salaries for statistical jobs would be, too. And there were certain personal appeals to life in a larger city: expanded professional networks, greater opportunities to find friends with similar interests, a vibrant Jewish community, the option to spend less of my life driving…

Plot Twist!

Going into the fall I was pretty confident that I would only end up applying to JPSM. However, my bosses at IQS complicated the decision by making an appealing offer: attend school part-time either at University of Kentucky, Georgia Tech, or JPSM’s MPS online program, but take on a promotion involving greater responsibility and additional learning opportunities. I would need to decide on this offer before hearing back from any of the programs. IQS wanted me to take on the promotion as a first choice or not all, a condition which I understood and respected. So essentially my choice amounted to either applying to JPSM and leaving for DC, or taking on a new commitment at IQS and continuing to work while completing a part-time or remote master’s degree.

I spent several weeks wrestling with the decision. Fortunately, I was not alone. People I respected and loved generously offered an ear and advice to help me weigh it over. Izzy (my wise older sister) and Mauro (my old pal and self-appointed spiritual adviser), among others, were especially helpful. Nonetheless, I really struggled with the decision. Up until the last possible moment, I was unsure what I’d ultimately choose. In a scheduled meeting with my bosses Shawn and Christina in Shawn’s office, I needed to give them an answer. During my lunch break beforehand, I went for a long walk nearby and anxiously weighed the pros and cons. When I walked in the door for the meeting, I honestly didn’t know what I’d say. But when I sat down and Shawn asked what I’d decided, it simply came out: I truly appreciated the offer, but I was going to pursue full-time, in-person grad school and would carry on applying to JPSM. Shawn and Christina were disappointed but took it in stride, and to their great credit made it clear that they were still in my corner and would move forward in a way that would make this a win-win opportunity for all of us. They’d be able to leisurely bring on my replacement, and until grad school started we’d carry on working together on impactful client projects and interesting long-term development projects.

I was relieved to be done with the decision, but I couldn’t help but immediately feel saddened by it. I would miss IQS and the people I’d worked with there. Shawn was a mensch and a mentor who had deeply impacted the way I work and think. I’d had countless great conversations with him in that office. Brainstorming a study design, marveling over the “everything old is new again” nature of big data hype, talking philosophy, having hard conversations about tough conversations… And I’d greatly enjoyed working with the rest of the IQS team at the time. Joel, the thoughtful and curious-minded head of data collections who improbably also had a brief stint studying Arabic and delighted in exchanging GIFs of Mary Berry from the Great British Baking Show. Hayley, the bright analyst who shared an office with me and made me laugh throughout the day. Laquisha, the hilarious and refreshingly plainspoken star interviewer at the time. Christina, who’d been a mentor and helped me feel at home in Louisville. I’d miss them all.

A combination mensch and mentor: what you might call a “menschtor.” I wouldn’t call him that, but you might.

Learning from the decision

The difficult few weeks of making that decision also served as a final push to get me to do something about the chronic anxiety I’ve always lived with and had generally taken for granted. Like David Foster Wallace’s fish who doesn’t know what water is, I had spent nearly all of my life unaware that I lived in a constant state of needless anxiety and self-pressure. Fortunately, two people that year had kindly and independently suggested that I seemed constantly worried and excessively hard on myself, and perhaps I should consider seeing a therapist about it. The first was my boss Christina, and the second was Dr. Miller, the UofL professor in real analysis. They were both right.

That winter I started seeing a therapist, who diagnosed me with generalized anxiety disorder and started me on a course of cognitive behavioral therapy.¹ I came to see that so much of my incessant worrying was more a habit than a rational response to circumstances, and that so much of the constant pressure I felt was a sort of optical illusion in how I thought about myself and my place in the world. I learned that my tendency to procrastinate was a predictable consequence of perfectionism rather than a stubborn challenge that could be conquered with perfectionism. I came to appreciate the deep wisdom of my boss Shawn’s advice that perfection is a goal that is generally both unattainable and not worth the price. I resolved to start treating “good enough” and “done” as the general goal in daily life, and to gently remind myself that this attitude is not some sort of shameful concession to mediocrity but rather a healthy, clear-eyed way of dealing with the many goals I have for my life and their competing demands on my limited time.

And that’s why this blog exists: to practice working on something until it’s “good enough”; maybe a little long-winded, almost certainly too personal, but good enough.

Footnotes

1. This wasn’t the first time I had seen a therapist. My last year of college, I saw a school therapist at the mental health center when I was dealing with situational depression largely related to serious problems with my parents’ health and their financial straits, which included bankruptcy, foreclosure, and other challenges. I don’t think the college therapist was able to recognize my chronic anxiety for what it was, given the family issues I was dealing with. But on top of that, constant worry and perfectionism is unremarkable at Amherst College, an elitist institution where students and staff rationalize their lucky position in the social hierarchy by imposing grueling high standards on their professional and personal lives.