NYU MS Data Science vs Columbia MA QMSS

[ceroper]

Hi folks,

I've been accepted to the following programs:

MS in Data Science from NYU

MA in Quantitative Methods in the Social Sciences at Columbia (want to pursue the data science concentration)

MS in Computational Analytics and Public Policy at University of Chicago

Master of Information Systems Management with a concentration in Business Intelligence and Data Analytics at Carnegie Mellon

So exciting to have so many options! I think I've probably narrowed it down to Columbia and NYU, because those programs are most narrowly focused on the data analysis stuff I find most interesting, but feel free to persuade me otherwise! I worked at a marketing consulting firm in the analytics department for two years, and then I worked at GlobalGiving, a nonprofit crowdfunding platform for a year and a half. For undergrad, I studied Economics and Professional Writing at Carnegie Mellon University. I've taken some programming here and there, but doing business intelligence work using SQL and Excel has been by bread and butter, so any of these programs would stretch my technical skills.

I'm not completely sure what I want to do when I graduate, but broadly speaking, I want to be a data scientist in the social impact space. I could see myself as a data scientist at a crowdfunding site and/or a number cruncher at an international institution like the UN (Global Pulse) or the World Bank.

Which of these programs would get me where I want to go? If you can't compare, what could you say about any of these programs? I'm curious about the reputation and name recognition of each. If you graduated from any of these programs, how did you like your experience? For NYU and Columbia especially, these span multiple departments and are really flexible so I'm curious if it's hard to feel like you're part of a community, because everyone is so dispersed. These are also fairly short - Columbia's QMSS is only 10 classes and NYU's MS in Data Science is only 12 classes. Did you feel like the program stands alone and did you have enough time and classes to learn what you needed?

Thanks for your thoughts!

 

[juilletmercredi]

This is just my opinion. I think both programs would get you to jobs that you would enjoy, given your interests. I would lean towards the Columbia program if your main emphasis/interest is in the social sciences part, you really want to learn more about integrating social science interpretations/applications to the quantitative analysis you're building, and you feel like you already have a pretty solid foundation in statistics and CS (or are confident you can build these skills on your own). I'd lean more towards the NYU program if you want to focus more hardcore on the data science part, including more theoretical work in statistics and probability; if you think you may be interested in private sector work maybe in the future, or want the flexibility of choosing; or you feel like your stats chops could use some improvement. For context, I went to Columbia and heavily considered the QMSS program myself before deciding instead to get a PhD in a different but related field. (IF there was a QMSS PhD program, I would've been all over that!)

The QMSS program seems really focused on social sciences who want to develop their statistics and research methods skills to the point where they could be The Research/Stats Person on a team of social scientists - but not really a statistician or data scientist. It's kind of evident in the alumni profiles they have - an associate economist at a public interest business think tank; an analyst at the Bureau of Economic Analysis; a senior fund analyst at an investment bank; an analyst in NYC's social services department. Essentially it prepares you for roles in which you'll be doing a lot of quantitative analysis in a social science context - hence the name. Could you go be a number cruncher at the World Bank or UN? Maybe, especially if you already have SQL and analysis experience. Could you be a data scientist at a crowdfunding site or a hospital or another social services agency? Eh. Unless you developed a significantly broader set of coursework during the MA program - to include more programming and coursework in the statistics department focused on big data, maybe some in the CS or OR departments focused on machine learning and data mining - they might go with a statistics MA or CS MA over you.*

NYU's program is more focused on building competency in statistics (mostly applied, although I see some theoretical coursework in there) and some computer science. The focus is much more on coursework that's directly relevant to manipulating and analyzing big data. The choices of electives are more spread out in field that are directly relevant to that, too - general CS, large-scale computation, bioinformatics, mathematical finance, business analytics. But the only social sciences classes that are approved are ones that are very directly relevant to analysis (like econometrics for economics, measurement theory in psychology, quantitative analysis in political science and sociology). There's less of that "for the social sciences" woven in, and I'd be willing to bet a lot of money that most of the applied examples in your coursework will be primarily business, computer science, or maybe biomedical/biotechnology related. Very few will be social science-related, most likely.

I'd also imagine that the peer groups are much different: the NYU students are going to be more focused on finding data science jobs in industry (probably primarily tech), whereas the QMSS students will be all over the map but probably much more likely to want to find public sector or public interest work post-graduation. That makes a difference - not only in encouragement and social capital amongst your peers, but in the kinds of speakers and alumni and recruiting the program brings to campus as well as the networking that you do after you graduate.

Of course, with the flexibility of QMSS you could probably add more coursework from the CS, OR, and stats departments at Columbia to round out your program. One thing I have found about Columbia is that all of their departments tend to be relatively flexible when it comes to fulfilling requirements (undergrad and grad level, which is what makes me think it's sort of a university trait and not just limited to certain programs). I'd ask about that ahead of time. For example, the QMSS website lists electives like machine learning, applied data mining, and advanced data analysis. I happen to know that there are more in other departments (and in those) that are relevant to data science and that Columbia has a lot of support on campus for data science: despite the Institute being relatively new, there have been professors doing research on that for quite some time. (However, I'm pretty sure that NYU has at least a comparable level of support for it).

So, the tl;dr is that I think you can't really go wrong, but it kind of depends on which way you lean. More public interest/non-profit-y, and contentedness with being primarily a quantitative social scientist? QMSS might be better. Want a mix, but really want to develop your chops to be a data scientist or statistician? Then NYU might be the ticket.

[ceroper]

Thanks! This was really helpful and thorough! I've got a lot of thinking to do, but it's great to have your perspective, especially on the differences between the peer groups.

[GM1993]

NYU Data Science is a more selective and competitive program than Columbia. Quantitative Methods in Columbia could be recognized as a "Step Stone" for further study, like preparation for PHD, which will give you more opportunities in research area.

[Jessie Z]

I think NYU is better.

[rack_attack124]

is anyone on here going to QMSS? I'm waitlisted currently. Fin aid emailed me back and said cost of attendance is $86,951. Anyone know if this is worth it? More than likely, I would use the degree to find a job and not go into PhD programs. Trying to find what kind of salary could be expected after QMSS.

[juilletmercredi]

FWIW I don't think selectivity of a program, or competitiveness, really matters. Those are inputs. What you are interested in is outputs. A program could admit 50% of its applicants because it's a small niche program with a self-selecting group of applicants, but 100% of its graduates get jobs within 3 months of graduation and they all make good money. So IMO, ignore "selectivity", which is usually measured in terms of acceptance rate. Ask instead about placement rate and starting salary and what people do after. They sometimes overlap but they are not 100% correlated.

As for the salaries - I think the issue here is that because there are so many things you can do with QMSS, there is no one "expected" starting salary. You could end up as an analyst at a small NYC nonprofit making $55K or you could end up as an associate economist at a large multinational making $75K. It also depends a lot on the skills you come into the program with and the concentration you take - someone who knew how to program ahead of time and took the data science concentration might have higher-paying job opportunities than someone who is concentrating more on very applied social science type stuff. With a one-year program I'm betting a lot of job searching is going to be pretty dependent on already held competencies.

[rack_attack124]

20 hours ago, juilletmercredi said:

FWIW I don't think selectivity of a program, or competitiveness, really matters. Those are inputs. What you are interested in is outputs. A program could admit 50% of its applicants because it's a small niche program with a self-selecting group of applicants, but 100% of its graduates get jobs within 3 months of graduation and they all make good money. So IMO, ignore "selectivity", which is usually measured in terms of acceptance rate. Ask instead about placement rate and starting salary and what people do after. They sometimes overlap but they are not 100% correlated.

As for the salaries - I think the issue here is that because there are so many things you can do with QMSS, there is no one "expected" starting salary. You could end up as an analyst at a small NYC nonprofit making $55K or you could end up as an associate economist at a large multinational making $75K. It also depends a lot on the skills you come into the program with and the concentration you take - someone who knew how to program ahead of time and took the data science concentration might have higher-paying job opportunities than someone who is concentrating more on very applied social science type stuff. With a one-year program I'm betting a lot of job searching is going to be pretty dependent on already held competencies.

I agree with you and the website said that most people in QMSS already had work experience but a few came right after undergrad. I don't have much work experience and I also didn't just graduate. I think for the amount of debt I would have to take, then the salary would have to be pretty high! I don't want to be paying loans for the rest of my life!

[tiy005]

Just curious: where did you end up going?