Curious of possibility...To stats or not to stats

[ReynoldWoodcock]

Hi All…I’m a long time reader yet reluctant writer with a question. As the 2018 cycle comes to an end (congrats to all by the way) I have ramped up my own preparations for the fall of 2019. My question resides in whether you all think it necessary I take a statistics course to boost my chances? Whether it will be a lost cause regardless or prove needless.

I’m currently serving in the Peace Corps as an English teacher and teacher trainer at an Islamic middle school in Indonesia (I’m proficient in Bahasa Indonesia). I’ve begun to write my SOP’s/reach out to potential recommenders. It’s early, but I have oodles of time and want to make sure everything is done as well as possible. Prior to my time in Indonesia I was working for a rep. in the U.S. House of Reps for a year and a half as a recent college grad (think Staff Assistant role). I have a history of volunteering around the D.C. area as well as the more rural setting of the university I attended. While my resume is good on this front, I know it doesn’t scream out to admissions offices at the programs I hope to attend. I have a very specific path I hope to take following grad school but would rather not get into it. Know I will highlight these plans in my SOP’s and curb them to the specific schools I apply (Southeast Asia plays a role and I hope to ultimately take the foreign service exam in time).

So here’s the rub… my resume falls squarely on that pesky GPA from university (I’ll be 4 years out by the time I apply) and my years of immaturity (a few other factors I will make known…thankfully it has an upward trend). My GPA ended up a bit above the 2.9 mark. During my time in undergrad I did fine enough in my econ classes (think B’s) but took one statistics course for which I got a solid C. I have taken the GRE and pulled off a 160 on the quant. with a 165 on the verbal. I hope the GRE may help offset school’s worries since they take a holistic approach but I just never know.

So my question resides in peoples’ thoughts on me taking an online stats. course for credit to try and further separate myself from the underwhelming record I amassed as an undergrad? Whether it will help boost my chances at the schools below or whether I should rethink of applying to certain schools with the GPA I’m coming in with (particular programs fit me more than others for various reasons but I thought I’d keep it general for the sake of this post). Some schools may be a big ole reach but thought I’d try nonetheless. Thanks for any advice and I hope this can help others in the process!

1. Fletcher
2. SAIS
3. HKS
4. SIPA
5. MSFS
6. UChicago
7. Duke
8. Carnegie Mellon

[Damis]

I think your GRE score more than makes up for the GPA and C in stats. I've also seen folks with lower GPAs get into some of the schools you listed.

Given you've already taken stats, you're more so asking whether or not you should retake the class? A normal 101 stats course? No. I'd try to take a more difficult course...such as Econometrics 101 or a higher level stats course. If you have the time to study and are wanting to show the various AdComms that you have the ability, that would be more effective. Given it'd be just the one course, I think you can do it!

I'd advise taking a gander at what you'd be studying, however, because it may be too broad a leap. Check out this MIT Econometrics course. If it looks feasible, do it. If not, then maybe a higher level macro would do the trick.

[79ajjohnson]

My two cents.

MPP/MPA can be really heavy in quant, especially stats and economics (Harris told me this on my visit). I don't think they'll fault you too much if you can come up with a solid explanation as to why your GPA was low and why it shouldn't affect your admission decision (that's what I did).

I took two stats courses, one calc, one finite, and at least five other quant heavy courses in undergrad, but what a lot of these programs are is stats and econ. Learn how to run regressions (it really isn't too bad, plus you scored well on the quant section of the GRE), how to data mine, and how to interpret results and draw conclusions. If you have written any papers for stats or any quantitative course, use that for writing samples. I personally used two stock write-up's from my investing courses and a 20-page paper I had for my stats: decision making course. 

You seem to have a lot of experience in the field too which can also really help make up for your low GPA and you have strong GRE scores. Spend a lot of time on your essay, as I think that was what pushed me over the edge.

[3dender]

It depends on the program you want.  For most of the international-oriented programs at the top of your list, an extra stats course is probably not going to move the needle (especially since they can see your solid aptitude from the GRE).  It might impress Harris but I can almost guarantee you that Duke would not care either way (I'm a 1st year at Sanford).  Lots of folks came to my program with little to no stats experience, and us students are even pushing them to deemphasize it further.  Your general (non-statistics) GPA will be the biggest factor working against you for HKS unfortunately.

The truth of the field these days (as explained to me by various soon-graduating 2nd years) is that for the highly technical quant work you'll see at places like Harris, most firms would rather hire PhD statisticians than MPPs.  Knowing that stuff is just not necessary for the vast majority of policy/research analysts anymore.  Yes knowing the basics of R/Stata is an important skill to develop, but administration assumes no knowledge with these programs for a huge part of the cohort.

[yoh_rrg]

10 hours ago, ReynoldWoodcock said:

 

1. Fletcher
2. SAIS
3. HKS
4. SIPA
5. MSFS
6. UChicago
7. Duke
8. Carnegie Mellon

It might be worthwhile to try to figure out the quant requirements for each school. I say this primarily because I see CMU on your list, and they made passing stats with "at least a B" a condition of my acceptance. For some of the schools taking stats (or a more difficult quant course) might simply demonstrate how serious you are about returning to grad school, while for others it might be a requirement for enrollment. You just need to read up on these schools and their "eligibility" requirements to see if you need it. Naturally it wouldn't hurt to take a course, but since they're pretty expensive I would only take one if necessary for any programs you've selected.

[yellina122]

I would strongly recommend taking an econometrics or stats course! I took econ through Erie Community College online and it was both a cake-walk and definitely significantly contributed to my getting into my top school (SIPA). I had actually talked with a couple admissions reps from SIPA beforehand and they expressed that despite getting 159 on the quant section, because I didn't have a strong history of coursework for economics, I should take a class if I can.

 

It's kind of costly (~$1k if I remember correctly), but can be worth it!

[ReynoldWoodcock]

Thank you all, this is some great advice. I've been wavering between all these options. Whether to check out an econ course such as econometrics or bite the bullet, not spend the money and see where my application as a whole takes me without the additional course.

 

I appreciate all of you responses!

[wittgensteinsbladder]

On 3/21/2018 at 5:45 AM, 3dender said:

The truth of the field these days (as explained to me by various soon-graduating 2nd years) is that for the highly technical quant work you'll see at places like Harris, most firms would rather hire PhD statisticians than MPPs.  Knowing that stuff is just not necessary for the vast majority of policy/research analysts anymore.  Yes knowing the basics of R/Stata is an important skill to develop, but administration assumes no knowledge with these programs for a huge part of the cohort.

What does the field look for instead in your view? I.e. What sort of classes are you pushing to replace stats coursework at Duke?

[3dender]

As with most things, it depends.  My sense is that for probably 2/3 of my class, our future employers will merely want us to have significant experience heading up client-facing projects, which Duke absolutely provides.  They'll also want to see you passionate about their specific field, which is taken care of by the decent variety of subject-area classes.  Everyone comes out of here with at least decent research and analysis skills, so there's nothing there that is really going to separate one candidate from another (again, for that 2/3 faction -- the heavy quant folks I'm sure will land wherever they want in their specialty).

I've personally been okay with the stats coursework.  The first semester (for remedial folks like me) is sort of the basics that takes us up through regression analysis via Stata and also study validity. The 2nd semester focuses on program evaluation which I think is an important skill, also teaching study designs beyond RCT like difference-in-difference, regression discontinuity and instrumental variables.  What I'd change is adding in a stronger research interpretation perspective where we have to regularly judge the validity of academic work.

[Arturo Belano]

On 3/26/2018 at 1:11 PM, 3dender said:

I've personally been okay with the stats coursework.  The first semester (for remedial folks like me) is sort of the basics that takes us up through regression analysis via Stata and also study validity. The 2nd semester focuses on program evaluation which I think is an important skill, also teaching study designs beyond RCT like difference-in-difference, regression discontinuity and instrumental variables.  What I'd change is adding in a stronger research interpretation perspective where we have to regularly judge the validity of academic work.

Hi 3dender,

I am currently trying to decide between Chicago Harris and Duke Sanford for an MPP and I am trying to figure out how well prepared the Sanford program would leave me for a research heavy job/PhD. Harris is known for being quantitatively rigorous, but unfortunately they provided me with significantly less financial aid than Sanford. Do you know how research/quantitatively oriented the Sanford program can be? Are Sanford grads usually well equipped to pursue a PhD? From looking at the course listing, it seems like it is more focused on domestic US policy and issues like lobbying, government decision making and governance, so I hope there is room for more quantitative stuff.

Thanks a lot! 

[3dender]

3 hours ago, Arturo Belano said:

Hi 3dender,

I am currently trying to decide between Chicago Harris and Duke Sanford for an MPP and I am trying to figure out how well prepared the Sanford program would leave me for a research heavy job/PhD. Harris is known for being quantitatively rigorous, but unfortunately they provided me with significantly less financial aid than Sanford. Do you know how research/quantitatively oriented the Sanford program can be? Are Sanford grads usually well equipped to pursue a PhD? From looking at the course listing, it seems like it is more focused on domestic US policy and issues like lobbying, government decision making and governance, so I hope there is room for more quantitative stuff.

Thanks a lot! 

Hi Arturo. I too was deciding between Harris and Sanford.  At their Open Houses (which I encourage you to attend if you're able) my impression was that there wasn't very much difference between the amount of quant emphasis of each program, somewhat surprising since I too was under the impression that Harris was especially quant-heavy.  I think Sanford is consciously emulating that aspect of Harris.

You're right that Sanford has a relatively heavy social/domestic policy focus.  There's also a sizable NatSec contingent, and we're encouraged to take MIDP classes if we want as well.  If you have a solid stats background you would place into Advanced Econometrics from the outset.  The class of '18 only had to take one semester of stats for advanced students, but our year had to take two semesters (not sure if this will change back or not, several of my cohort complained about it).

I do know there are more than enough quant electives to fill your entire schedule if you so choose.  Your 1st year is pretty prescribed (4 core classes the first  semester, 3 the second, with most students choosing to take 5 classes in one of those semesters. . . equaling two stat courses (core) your first year and the possibility of two more through electives).  To give you a taste, these are the non-core stats courses offered next semester, which seems fairly typical: Public Expenditures Evaluation, Cost-Benefit Analysis, Public Finance in Developing Economies (MIDP), Public Budgeting and Financial Admin (MIDP), Policy Evaluation with Data, and Regional and Economic Social Analysis (I'm thinking of taking this one, which deals with how to find, search, clean and manipulate publicly available datasets).  

To answer your other question, yes Sanford grads are well equipped to pursue PhDs.  Several of my classmates have this goal, and professors and PhD students are more than happy to work with you on that goal if you ask them.

I hope this helps.  Let me know if you're going to be coming to the Open House next week -- I can hook you up with some of the advanced quant students.  Actually I can do that either way now that I think of it. . .

[coasts]

I had similar stats (college GPA, work experience) and decided to take an online stats course the fall while I applied to school (fun!). Obviously I don't know what the counterfactual would have been, but I ended up being fairly happy with my outcomes (especially considering that GPA...) and feel like the class must have helped ad comms feel more comfortable about my academic abilities. So I feel like it was a worthwhile thing to do. Plus, it doesn't hurt to brush up before returning to school. 

[chocolatecheesecake]

On 3/30/2018 at 4:37 AM, 3dender said:

To give you a taste, these are the non-core stats courses offered next semester, which seems fairly typical: Public Expenditures Evaluation, Cost-Benefit Analysis, Public Finance in Developing Economies (MIDP), Public Budgeting and Financial Admin (MIDP), Policy Evaluation with Data, and Regional and Economic Social Analysis (I'm thinking of taking this one, which deals with how to find, search, clean and manipulate publicly available datasets).  

On a side note, @3dender, I hope you're going to sign up for Public Expenditures. Bob Conrad is one of the most amazing human beings alive. His class is incredibly funny, useful, challenging, practical, and fun; even when he was called to NYC to help the government of Ukraine patch their budget back together, he insisted on setting up an A/V link so he could Skype his class to us live. This is the kind of knowledgeable and dedicated professor you need in graduate school. (Even if his accent is the thickest Southern drawl I've ever heard.)

[chocolatecheesecake]

Also @ReynoldWoodcock, I have two friends who were in the Peace Corps in Indonesia and attended Duke's Sanford School with me. One of them is back in Jakarta working at J-PAL, actually, and another is in DC with one of the big consulting firms. If you're curious and want to ask them some questions, I can definitely hook you up with them. Just PM me your information. 

[juanlisa]

I was in a similar situation. Overscheduling myself in undergrad led to a C+ in calculus and a B- in intro stats. Taking online math courses in calc 2 and linear algebra (and doing well) more effectively demonstrated my capacity for math. But possibly more important: the courses made me less anxious about applying. For what its worth, those courses and a decent GRE score were enough to get me into the analytics track at CMU and the data science track at Georgetown.