Thanks to David Wickland MIA ’19 for this post in response to a topic submitted by Nicole H. Submit your idea for a blog post here.

Taking a more quantitative focus at SIPA can mean a lot of different things: There’s the underlying conceptualization of quantitative analysis taught in Quant I and Quant II, the more direct applications covered in Evaluation and Economic Development classes, the academic literature analysis of the various Quant III classes, and the programming focus of others.

They’re all important. They’re all interesting. You might not want or need to do all of them, but there’s a lot to choose from at Columbia.

I’m a 2019 SIPA grad who obtained an MIA with a Concentration in Economic and Political Development and a Specialization in Advanced Policy and Economic Analysis. I studied electrical engineering in undergrad and had worked as a data analyst prior to SIPA, so I came in with a decent quantitative background. But I had little-to-no knowledge of how quant could be applied in the social sciences.

One of my goals at SIPA was to figure out how to go about using any of this stuff, and taking as many quantitative courses as possible seemed like a good way to explore different applications.

Quant I

I never took Quant I, so I’m going to gloss over it a bit, apart from noting that regardless of how you feel about it, don’t let it play too much into your decisions on other quantitative classes. It covers a lot of ground and can be a bit overwhelming, but the later courses tend to take a milder pace and help drive home the topics covered in the first semester.

Quant II

This brings me to Quant II, which I cannot recommend enough. This is probably the second quant class most people will take (although exactly what constitutes a quant class is debatable). Quant II essentially picks up where Quant I leaves off, delving into the most widely used regression methodologies to give enough understanding to follow most papers and studies one would come across. I don’t want to claim that everyone loves Quant II, but a lot of people who thought they would hate it wound up loving it, and I think it’s probably the best course to judge if this is something you like. It gives a more practical understanding of the material and helps reaffirm everything in Quant I.

(Just to note, there are currently two Quant II professors, Alan Yang and Cristian Kiki Pop-Eleches. They’re both great, and their classes are structured slightly differently. In Alan Yang’s, the last month is spent on a data analysis project that helps ground some of methods in more practical usage, while Kiki’s ends by covering some additional methodologies which are also useful. If you take Alan Yang’s, the skipped methodologies are covered in Applied Econometrics and Economics of Education Policy;  if you take Kiki’s, Harold Stolper’s Data Analysis for Policy Research and Program Evaluation is essentially a full semester version of the data project you would have done, so neither is entirely a missed opportunity.)

Quant III

While the Quant I – Quant II track has clear continuity, classes after these focus more topically and can mostly be taken in whichever order one likes. The term “Quant III” gets thrown around a lot, and it refers to a group of topically different classes which require Quant II, not a single specific class. In no particular order, these are some thoughts on the Quant III classes available:

  • Applied Econometrics: This covers a lot of the loose ends and more in-depth examinations coming out of Quant II, and is probably the most direct follow-up to that class. It is very technical compared to Quant II and less immediately practical. Quantitative Methods in Program Evaluation and Policy Research, which was not offered during my second year, is supposedly similar but more applied.
  • Economics of Education Policy: If you are interested in education you will love this class. It explores different aspects of education and the research surrounding them, with general open discussion of the papers, their relative merits, and their implications. Very highly recommended.
  • Time Series Analysis: This is perhaps the most technical Quant III class, and it has a fairly narrow focus on financial markets and predictions. For students with good quantitative and programming skills and are interested in how markets can be tracked and the underlying principles of time series’, this class is for you. If that’s not your cup of tea try one of the other classes instead. (Note: This class is taught in R, and is the only class at SIPA to do so to my knowledge. The basics are explained and the coding is not particularly intensive, but it can make things difficult. At the same time R is wonderful and everyone should learn R.)
  • Data Analysis for Policy Research and Program Evaluation: Whether or not this is a Quant III class is debated, but it does require Quant II and covers quantitative material, so it’s at least related. Full disclosure, I never took this class, but I generally heard positive things about it. The course is a semester long data analysis project, and works to build a deeper understanding of STATA both in the data analysis and data visualization fronts. I generally heard excellent things about it, and would recommend for anyone who wants to learn more applied STATA.

Thoughts on other Quant classes of interest:

  • Computing in Context: This was good introduction to Python as a language. The applications aren’t particularly quant-oriented, but if you’re looking to learn Python this is probably the best way to go about it.
  • Program Evaluation and Design: Not a quant class per se, but I feel that most quant classes at SIPA are focused on research and evaluatory studies. This class (which I did not take but have heard great things about) can help fill in more around how data was collected and why that specific question was asked or that specific information was gathered.
  • Machine Learning for Social Sciences: Taught in Python, this Quantitative Methods in the Social Sciences (“QMSS”)** class goes into the fundamentals of machine learning and its applications. For any SIPA students interested in ML or AI, this is probably one of the most directly applicable courses available, although QMSS students get priority and it tends to fill quickly.
  • Data Mining for Social Science: Taught in R, this QMSS course is the main Columbia class on data mining and it’s supposedly fairly good as an introduction. This is another class I never took, but what I heard from other SIPA students is that it was interesting, though not particularly in depth.
  • Statistical Computing with SAS: This is a Mailman School of Public Health course on SAS. I knew one person who took this and they seemed satisfied. It sounds similar to Computing in Context except for SAS and with more of a public health focus. SAS as a language isn’t nearly as common as STATA/R/Python, but it’s still useful to know. It’s also quite different from the other stats languages and can be harder to learn on your own.
  • Research Techniques and Applications in Health Services Administration: This is somewhat similar in design to Economics of Education Policy except it is at Mailman, a bit less technical, and focused on Public Health. If health is a particular interest area and you want to know more about the quantitative studies surrounding different aspects of it, definitely try to get into this.

If you’re interested in SIPA’s quantitative program, I recommend researching and asking around about the courses you want to take. For example, talk to current students who may have taken the courses you’re interested in, speak with faculty members such as Kiki and Yang, and take a look at the course evaluations on a specific class as well as old syllabi.

I talked a lot with Kiki about what courses I was looking for, and he gave me a holistic view of the Quantitative program and an overview of the course’s strengths and weaknesses. I found his guidance valuable, and coupled with my research on the courses I wanted to take, I was able to craft the quantitative experience I was looking for at SIPA.

**QMSS is housed in the Graduate School of Arts and Sciences. SIPA students can take courses through QMSS by cross-registering, as well as obtain a dual-degree through its program. For more information about QMSS please visit their website here. For information about Columbia Dual-Degrees, visit our website here.