The GC’s Quantitative Research Consulting Center
The Quantitative Research Consulting Center (QRCC) provides resources for statistical support in quantitative and empirical research. It complements existing statistics coursework by bridging the gap between the classroom and implementation in researchers’ own work. The new Center provides support for MA students, PhD students, post-docs, and faculty in any CUNY GC program at any stage of research. Support is provided through a consulting model that advises students on choices of statistical methods and packages and gives tailored help and advice in implementing analysis.
In terms of statistical methods, the Center helps GC students determine and interpret appropriate statistical tests: including parametric and nonparametric fundamental tests (such as t-tests, correlations, ANOVAs, regressions, and chi-squared), as well as advanced tests (such as factor analysis, principal component analysis, structural equation modeling, mixed and multilevel modeling, cluster analysis, bootstrapping, and Bayesian statistics). It also helps you select statistical packages (such as R, SPSS, Python, MATLAB, SAS, HLM, and STATA) and understand their output.
If all that sounds like a foreign language, fear not! I sat down with Christen Madsen, the head of the new Center, to get a better sense of how and why you can use statistical consulting in your dissertation research–even if you don’t consider yourself a quant.
Christen co-hosted an event with our office, “Writing an Article in the Quantitative Social Sciences,” on March 1, 2018.
Q&A with Christen Madsen
Q: What kinds of services do you provide for students?
A: We do statistical design and research consulting. When you’re doing a project for your dissertation, or any kind of research, there are many steps involved. You have to lay out the way you’re going to collect your data, you have to think in advance of what it’s going to look like when you analyze it. That’s the research design part of it. It’s all about helping you plan the beginning of a project. Then you go off and get your data, and then you come back, and then we talk about how you analyze it. So that’s the bird’s eye view of statistical consulting.
All of that goes on at the back end. At the front end, the statistical design and research consulting takes the form of “how can you operationalize your research questions?” It’s about helping you figure out how you’re going to answer those questions. What data do you need to answer those questions? Once you get the data, it’s about being deliberate and intentional about what it’s going to look like when you analyze it. You have to think about all these things in advance to make sure you know what you’re doing. You should know what you plan to do before you start looking at numbers. It’s not just about deciding what question your trying to prove, it’s also about deciding what “proving” the question would even look like. Is it statistically significant differences in terms of T-tests? Is it about understanding relationships or correlations between things? Is it just simple counts (for example, saying there’s 100 different times something happens here, and zero times it happens there)?
From there, we can also help with Institutional Research Board (IRB) documentation. And it’s not just for quantitative research or the so-called bench sciences. Let’s say someone comes in from the English department. The consultation would start by asking them “What is your research question?” Then we’d be thinking about ways you can turn those research questions into numbers. That’s what we mean by operationalizing. It’s figuring out ‘What can become a number?’ The important part is that designing your study, and how you can operationalize it, starts at the beginning—before you’ve gone and collected your data. It has to start at the beginning to figure out how you can translate your qualitative data into something numeric that helps you understand broader trends and patterns. Learning to do some thinking up front, before you start your research project, is useful even if you don’t do statistics.
Q: What departments are people coming to you from?
A: Archeology, physical anthropology, chemistry, business, political science, all the psychologies (developmental, clinical, educational, psych and law, neuroscience), urban education, linguistics, and so on. Faculty also come in, not just students! For example, faculty may want to understand how they can help their students develop the right kind of statistical approach to their methods.
The basic process for a consultation is that I talk to people, refine their statistical plans, or help them develop their statistical plans. When it comes to running the statistics, I help people program, interpret the meaning of the results, and then write up their results. I think what’s great is that this Center is solving big need in the CUNY community. I mean, this is what I was doing in the Linguistics department for three years. When we created this center it was an extension of what we’d already been doing for years. It’s not something I set out to do. My first client was someone I was already working with.
Q: People often find statistics intimidating. Do you have a philosophical outlook on how people might start to integrate statistics into research projects that are qualitative?
A: When someone says they do statistics for fun, it’s generally a conversation-killer. People just get uncomfortable. And I think that’s for a good reason. For a lot of people in the qualitative sciences, statistics is their most painful class. I think this goes back to the fact that teaching statistics is difficult. Even when you take a class in statistics or coding, you may not feel confident that you can do it outside the class context. Teaching statistics is easier when you start with the basics; with an understanding of how you’re going to use a quantitative analysis, and when you would use it. Why would you use it? And what does it say? But many people feel that statistics isn’t explained in a way that has any meaning for them, so they file it away in the category of things you learned growing up that don’t really mean anything.
Statistics is a way of viewing the world. I feel like my role is to help people with their research design so they can start to view the world numerically—to coach them through that. What in your world could you turn into numbers? If you’re only talking to other people in the humanities, then you’re never going to get that quantitative angle. So it’s important to talk to someone who can help you translate it.
Q: From the professional development angle, what doors might learning statistics open that might not otherwise be available?
A: For someone who’s on a non-academic track, the simple answer is that it’s one more skill. One of the challenges with leaving academia is that, while there are skills that you gather over the course of working on a PhD, oftentimes you have to spend lots of time thinking about what those skills are—worrying about how to massage them, and how to leverage them, and re-frame them for a different field. But if you throw a chapter on statistics into your dissertation, that’s something you don’t even need to work on translating. You can just say, “I did this, I know statistics.
In the same way that talking about statistics makes people feel uncomfortable, it also impresses them. For better or worse, people are impressed by numbers—even academics are. Your dissertation is going to be long and hard anyway, so you might as well learn something different and useful while you’re at it! This process of using numbers and seeing numbers in the world is very useful, because academics respond to it, non-academics respond to it, and it’s a good way to translate your work across audiences. Also, when you’re working on your resume, everyone says you have to add numbers—whether it’s outcomes, outputs, results, or whatever it might be—because people simply respond better to it. I think anyone can benefit from learning statistics. Being on the job market is stressful for anyone. But knowing that I have this skillset gives me more comfort and confidence to face new career opportunities. It opens up a whole new world.
Q: How can students make an appointment with you?
A: Email me! You can reach me at firstname.lastname@example.org. Appointments last about 45 minutes, during which we talk about your goals. It doesn’t matter what state your questions are in. You don’t necessarily have to have your research question or your stats plan thought out before you come in. Maybe you know you need to do something, you just don’t know what it is. Those nebulous questions that you know exist, but you don’t know what to do about it or where to start, that is where I can help. I can also help people learn how to talk about statistics. So it’s fine if you’re not clear yet. Sometimes you just need someone to talk to about your idea and to be told that it’s not crazy.
Q: Finally, do you have any recommendations for external resources that people can use to get acquainted with statistics?
A: You can check out Nate Silver’s book The Signal and the Noise. I liked it because it’s helpful for thinking about Bayesian statistics. It’s a great book for learning how to think probabilistically. It’s all about building a model of the world—and that’s what statistics is. And when you’re answering a research question for your dissertation you’re essentially building a model—an artificial model—so you can see how things pattern within that model, which allows you to make extensions and inferences about the world itself.
There’s also The Cartoon Guide to Statistics—I liked it, it’s a fun book, and it doesn’t get into the fancy statistics. It spends a lot of time getting you acquainted with what probability is, and what sample size is, these kinds of basic concepts. Overall, though, I think the best thing to do is to read what people do in your field. Even if you don’t understand it, talk to people who do, and then just try it out for yourself. It doesn’t mean anything until you do it yourself.