Win condition

It takes 20 minutes to compile NoisePage every time I switch branches and I cannot do much on my laptop while it is compiling, so I've been having a lot of time to think lately. I've specifically been thinking about the oft-repeated advice of "you're in CMU SCS! take less classes, do more research!" in the context of graduate school applications. Research is what matters, not grades. Do grades matter? Think bigger! This advice is good, common enough to be mainstream, and widely-accepted by CMU SCS undergraduates. So why don't we take it?

Often times, even if a student wants to do research, wants to go to graduate school one hundred percent, and wants to take less classes, they may not. They might even feel that they cannot. This isn't even a discussion about whether graduate school is a good idea, we're taking that as a given. Instead, this is a discussion of my understanding as to why this phenomenon happens, based on my conversations with other undergrads and my own personal experience.

Metrics under consideration:

• Uncertainty, the root of the issue. More on this later.
• Finances, a secondary factor.
  • As of today, one semester of tuition costs USD 28,780.
  • Suppose you take a standard 9-unit course. If you take four courses, each course costs you USD 7,195. If you take five courses, each course costs you USD 5,756.
  • This ignores the additional opportunities that taking more courses opens up, e.g., another major, another minor, having room to fail and retake courses, having room for graduate courses, being able to take less courses when it matters.
  • Furthermore, if you're taking five courses, you can always designate one as "this course interests me the least, I'll focus on doing well in my other courses and my grades will average out", which is nice for your own mental sanity and sense of control. If you take four courses, you don't really have this luxury because that's the standard for remaining on-track to graduation.
  • Taking fewer courses makes each course substantially more expensive, in both financial and opportunity cost.
    • The people who say that additional majors or minors don't matter have never dealt with a paper-pushing society before.
    • The people who want to abolish grades are even more interesting when you spot them in the wild. The least-worst option... Anyways, back on topic.
• Area, a secondary factor.
  • "Taking one or two carefully selected graduate classes ... will often lead you to find interesting research problems" (document linked earlier).
  • But how do you carefully select these classes, and what if your selection was "wrong"?
    • You need three letters of recommendation, and the letters need to tell a somewhat coherent story. You don't want to sound wishy-washy in your SOP.
    • Human nature dictates that you'll click better with some professors and with some subject areas than with others.
    • If you highly focus your efforts in one class and end up being a B-grade student with few results to show for it, or if you just feel awkward around specific professors, then was it really worth it?
    • Not everyone can afford to do a masters program to make up for "wrong" choices. If I had to do a masters, I would not have considered research again for the next two decades.
  • Taking fewer courses only works if you know which courses to take to begin with. Otherwise, the "safer" approach is to shotgun courses until you can Texas sharpshooter fallacy your SOP.
    • Really good advice that I got from Andy: look for courses that cover recent research.
    • Really good advice that I got from peers: work with professors who bother learning who their students are. We have professors who didn't know who their TAs were, going into finals proctoring.
    • Personal advice: find a class with a final project that lets you explore related topics. It gives you something to talk about in grad school interviews, and it gives you a reason to bug the professor. Exams are of miniscule value here.
    • Personal advice, part two: look for the previous undergrads that a professor has worked with. Where did they go? Did they ever publish anything together?
• Imposter syndrome, a secondary factor.
  • Illustrated through anecdote.
  • "Yeah, I found concepts really hard, but I worked hard and made it through and so can you!"
    • I've heard this from many TAs and students. At surface-level, the message is inspiring.
    • What you don't usually hear is that many of those same students have participated in theory research, or taken college courses, etc., while in high school.
    • In giving advice, especially re: 112 vs 122, I make it a point to accompany every "122 wasn't too bad for me" with "but please note that I've been programming since high school".
    • In practice, I found the original message to be demoralizing, as did some other students. It makes you wonder "am I really cut out for this?" all too easily, without realizing that you were never at the same starting point to begin with.
  • The above example manifests even more strongly in the form of research engagement.
    • It "felt like there were institutions behind (student)" -- conversation with a friend. Some people walk in with a roadmap to college. From high school, into programs like Google CSSI, into maybe do research with X or maybe network with Y...
    • I met a friend because they wanted to join a research group as a first-year student. They already had useful university research experience coming in from high school.
    • I was reading Justine's twitter thread recently.
      • I got started in research because of Cortina's list of research opportunities. A little off-topic, but I wasn't really planning on doing it longer-term.
      • I stayed in research because by luck, I talked to Andy.
      • It still took a few months of Andy repeatedly telling me "you should hang out more in the lab!" before I realized that he really meant it.
      • It took me more time after that to be comfortable with the idea of measuring myself as a student with something other than grades.
    • Students also gossip, and a number of professors make casual career-crushing remarks that spread. I know another student where the professor was overheard saying "yeah, the undergrads I've worked with are pretty bad at doing research", and I've personally overheard "glad your student graduated, I didn't think that they would make it" from one professor to another. That's an oof if I've ever seen one.
  • Taking fewer courses is staking more of your reputation as a student on an intangible non-objective concept like research potential. It is much easier to get a transcript full of A's than it is to feel like you belong in a research group.
    • Even if you're sure that you want to go to grad school, it isn't clear whether your letters are strong enough -- maybe you'll have a better shot at trying to look like "got interested in research too late, but really strong grades and potential to be in the top 10 percent"?
    • If you're not sure that you want to go to grad school, then this is even more scary. The perception of grades being a bad indicator for overall competency is largely restricted to compsci programs in America. Between parental pressure, societal pressure, and conditioning from the past decade of schooling, it is easy to take comfort in a good report card.

And this brings us back to the original point: uncertainty. If you do not know that you will get into your desired graduate school for sure by focusing your efforts, then perhaps it is better to cast a wide net?

I think the best way that a professor can get an undergrad to commit to their joint research is to offer a similar commitment of "look, you'll definitely get into graduate school at X if you work with me". Which generally won't be done for multiple reasons: that's a liability that the university and professor doesn't want to take on, that's conditioned on past undergrad performance being indicative of future results, and because X would typically be the same home institution, that's not a good idea in the modern tenure-track world.

So, in the meantime, students will continue to hedge their bets, and advisors will continue to shake their heads.

---

Note: anecdotes from this post are not representative of the average CMU SCS experience. Many of the faculty are genuinely caring and kind. This is a post on why student interests may not match up with faculty interests, which itself would skew the perspective. Moreover, I'm incorporating multiple years of warnings and advice into a single post, so it will tend to have more cautionary stories than it normally would.