Can an MPP prepare you for a data analyst job at a private org?

[Nozistin]

Hi all,

Just thinking about career options here and seeing what folks think. Im wondering if an MPP can prepare one for data analyst/scientist jobs at a for-profit? Basically, make one competative with stat BAs or other applied social science majors? I applied only to HKS, GSPP, and WWS, the last two being very quant heavy schools. Im wondering what options that would open outside of public sector and non-profit jobs I guess. 

 

Thanks!

[pubpol101]

I would recommend Carnegie Mellon's MSPPM (MSc Public Policy and Management) program over all of those schools.

When looking at schools that consider themselves to be "quant heavy," there isn't much of a comparison. CMU's MSPPM program is housed in the same school as the information sciences college, and many public policy students take many quant courses that are difficult to find outside of a computer science/computational finance department. While typically schools might offer a few courses in econometrics, R, STATA, and other run-of-the-mill "quant" courses, there's much more at CMU. Many schools will even preface the syllabi of their quant courses with the words, "this is not a coding course," (as seen at HKS) when coding is taught at CMU and is exactly what you need for data science.

 https://www.heinz.cmu.edu/academic-resources/course-results/index.aspx . Just a few interesting ones are listed below:

• Programming R for Analytics
• Machine Learning for Problem Solving
• Text Analytics
• Big Data and Large-Scale Computing
• Data Focused Python
• Econometric Theory
• Management Science
• Decision Science and Multi-Criteria Decision Making
• Raster Geographic Information Systems

Your actual schedule works out so that you can take as many as six quarter-length courses in elective quantitative sciences in a single semester. You may be able to take even more depending on your exemptions and how heavy you want your load to be. So the availability of exotic and valuable quant courses is so broad that in a single semester, you can gain more of a quantitative background than a WWS or UCBerkeley (yes, I've checked their course lists) student might take three semesters or an entire two-year program to develop.

Data scientists are in demand, but the jobs are still competitive like any other. Given what I've seen on the course lists for HKS and WWS, I doubt their grads would be competitive for data science jobs when for-profits want incoming employees to know Python, R, SAS, ArcGIS, and/or whatever other tools are useful for the organization.

The primary deadline has passed, but you should still receive consideration if you apply very soon.

Edited January 12, 2017 by pubpol101

[ExponentialDecay]

Given that "data analyst at a private org" is an extremely broad category, yes, of course they can, and even the not-so-quantitative ones. Broadly speaking, I've met French majors who were data analysts at private orgs. There is a huge spectrum of organizations out there, and depending on the needs of a particular one, they may want someone who can build and test models, or they may want someone who can code bar charts in R.

Tangentially, what's with the random accounts putting up legit brochures for given schools in the comments? Is this the fabled phenomenon of schools paying unemployed graduates to shill their programs on grad school message boards? At least hire the clever ones, that know that most social science undergraduates these days are coming out with python, R, SAS, and ArcGis.

[Poli92]

I second @ExponentialDecay, it really depends on what you mean by "data analyst".  If you are leaning towards the data science "better at stats than any software engineer, better at software engineering than any statistician" direction then no, none of the programs you have applied to will give you adequate preparation.

However, there are lots of "data analysis" or "data journalism" roles in orgs of all types where you might be compiling summary stats and doing basic data viz. For those you really don't need anything more than rudimentary experience with a statistical programming suite or even Excel. If you want those jobs, you would be overqualified given your program list. 

If you want help refining your expectations, you should 1) start looking at the types of job postings you would apply to, and 2) go to some data analytics seminars or meet-ups in your area. 

[Nozistin]

On 1/14/2017 at 6:32 AM, ExponentialDecay said:

Given that "data analyst at a private org" is an extremely broad category, yes, of course they can, and even the not-so-quantitative ones. Broadly speaking, I've met French majors who were data analysts at private orgs. There is a huge spectrum of organizations out there, and depending on the needs of a particular one, they may want someone who can build and test models, or they may want someone who can code bar charts in R.

Tangentially, what's with the random accounts putting up legit brochures for given schools in the comments? Is this the fabled phenomenon of schools paying unemployed graduates to shill their programs on grad school message boards? At least hire the clever ones, that know that most social science undergraduates these days are coming out with python, R, SAS, and ArcGis.

Also to @Poli92 I guess more specifically what I am wondering is will these programs prepare me to work with large data sets in an analytical manner? Will I learn STATA and basic coding with SQL and/or Python? And how transferable the quanititiative skills are to areas out side of public policy analysis? I know these are generally broad questions with a simple answer being "if you don't want to public policy analysis then why do an MPP?" Well, I am primarily interested in public policy but am gaging what other doors this may help open and what skills I will be building. *Particularly important with the cost of HKS. Thanks for any responses in advance!

[Poli92]

The level of coding intensity will vary depending on the courses you choose to take, but frankly you shouldn't be going into an MPP program to learn how to code. You can learn to code in Python, R, SQL, and pretty much every other language (besides maybe SAS or some other high-end suites that keep a tight lock on training) online for free or cheap. Codeacademy, Udacity, Coursera, Udemy, O'Reilly Media, etc. offer a ton of great coding resources that will have broader offerings and deeper content than anything you will get from an MPP.  

In your courses you should focus on developing your expertise in your substantive area and developing your quantitative methodological competencies. In the summer or something try to get some sort of internship that will give you the opportunity to professionally demonstrate the coding chops you've developed on your own time. You'll also be amazed how quickly you'll pick up programming once you have the immersion experience of applying it in a professional environment. 

I guess all-in-all I wouldn't plan on shooting for a highly sophisticated data analyst role immediately after graduation. In all likelihood you'll be competing against people with CS or Stats backgrounds and they will be way more qualified than you. I would try to enter your substantive domain and then establish yourself as someone who demonstrates savvy with statistical programming, machine learning, data viz or whatever it is that floats your boat. 

(For the sake of disclosure, I work as an economist within a federal agency where I use a combination of SAS, SQL, and R for statistical research and application development)

[ExponentialDecay]

To add to what @Poli92 is saying, working with large datasets in an analytical manner doesn't mean knowing how to code them in Stata. Learning the software is trivial, and since which one of the many interchangeable suites you use is heavily organization or even client-dependent, being able to pick the coding up on the job is one of the realities of this type of work. Being able to analyze datasets means knowing the math. I would say that the amount of math in an MPA program is insufficient for a masters-level statistician job. Your strength would be your ability to couch whatever numbers you produce and the methodologies you use in your research context (e.g. agriculture, 4th industrial revolution, sub-Saharan Africa). Also, learn to write well.