Statistics MS Program Tiers

[PhysicsKid]

Hi everyone,

I hope you are all doing well during this stressful time of year. I was just wondering if there exists any type of thinking in terms of what the best masters programs in statistics are? I feel like it is even more difficult to rank Masters programs than PhD ones simply because of the variability in whether a Masters program is viewed as a kind of cash cow, a stepping stone to a higher degree, or a primarily professional degree. Personally, I would love to get some of your insights as to what the tiers of Masters programs are for someone who wants to get a PhD in Statistics, but I am certain that there are other people struggling with what to think of various masters programs and how to properly compare them.  I truly don't have any sort of insight into the matter, despite my constant attempts at googling; so, any insight would be greatly appreciated! 

 

A few example Masters Programs that I have no idea how to "order":

Oxford MSc Statistical Science, LSE MSc in Statistics (Research), Yale MSc Statistics and Data Science/Yale MA Statistics, University of Washington MSc Statistics - Advanced Methods and Data Analysis, Stanford MSc Statistics, Berkeley MSc Statistics, Columbia MSc Statistics, Chicago MSc Statistics

 

Thank you so much!

[supremum]

I know I'm going to get the wrath of all the alumni, but I actually think speaking to alumni in this case is the way to go. Each program has its quirks and feel to it, and oftentimes things will change pretty drastically in a year or two.

I can only speak to Yale, but if you get enough data points, you might be able to have a rough rank. First off, online has been a shitshow. But that's everyone, I imagine, except maybe those programs that have always had an online masters (?). I would defer if it were online. Pre-covid: small, tightknit group. I think if you're interested in going into industry after, there are better places (our career services are meh). I think our program is best for people looking to flirt with academia (many have taken the jump from masters to phd) or just get a very solid education.

[patient512]

7 hours ago, supremum said:

I think our program is best for people looking to flirt with academia

For those who are interested in continuing on to a PhD in statistics, would it be reasonable to rank the masters programs according to the USNWR Statistics ranking and omit the programs that do not offer research opportunities? I say this because I assume a one-year coursework/professional masters will not prepare one well for a PhD, nor add much substance to an applicant's profile.

By this reasoning the ranking would be roughly:

1. University of Chicago

2. University of Washington

3. Duke University

etc...

[supremum]

Masters->PhD jumps are easiest in the same department (the idea being that you're a hidden gem, and you don't come across that strong in your app, but once people meet you, they realize your worth). It's happened on a number of occasions in my department. The key assumption here is that you have to actually be an actual hidden gem (everyone thinks they are, obviously).

Rankings are definitely subjective, but let's just say that Duke being that high up discounts that construction in my (admittedly highly nonbayesian) eyes. But yes a good proxy for a good Masters->PhD program would be one with a good PhD program.

[PhysicsKid]

Hi all, thanks for the thoughtful responses! I guess the main impetus for my question was actually what @Stat Assistant Professor said in “Domestic Students Disadvantage“ which was that some highly ranked programs have very large, not necessarily great, Masters programs (cough Columbia cough) while some lower ranked programs have smaller competitive programs that are more akin to unfunded first and second years of PhD programs. I was trying to get a sense of whether there was any sort of consensus on which programs are best as preparation for a PhD? Would a one year program in the UK be better than one in the US? Is Stanford’s program more of an industry focused one or do people go from MSc to PhD there? There is precious little information online about many of these programs (even in terms of things as simple as size) so any information would be greatly appreciated. 

[patient512]

1 hour ago, PhysicsKid said:

Is Stanford’s program more of an industry focused one or do people go from MSc to PhD there?

I have a friend who recently graduated from this program. He told me that it's very difficult to find research opportunities with professors (both in and out of the stats department), because they prioritize working with PhD/Masters students in their respective departments. He said that most of the cohort is interested in entering industry after, but those who would like to apply to PhD programs tend to either do independent research, or assist PhD students. In both cases, however, this doesn't allow for one to form a working relationship with an advisor - which one may want to leverage during applications. With that being said, he and few other alumni have gone on to top PhD programs at Stanford and other great institutions, albeit not for stats.

Edited February 9, 2021 by patient512

[trynagetby]

15 hours ago, patient512 said:

I have a friend who recently graduated from this program. He told me that it's very difficult to find research opportunities with professors (both in and out of the stats department), because they prioritize working with PhD/Masters students in their respective departments. He said that most of the cohort is interested in entering industry after, but those who would like to apply to PhD programs tend to either do independent research, or assist PhD students. In both cases, however, this doesn't allow for one to form a working relationship with an advisor - which one may want to leverage during applications. With that being said, he and few other alumni have gone on to top PhD programs at Stanford and other great institutions, albeit not for stats.

I have a different experience. I was an undergraduate at Columbia and met a decent number of people (n = 3) in their first/second years in the Statistics department who graduated from Stanford's Stats masters program. They seemed to have a fine time finding research opportunities at Stanford as long as they could show genuine interest. Specifically taking some neuroscience classes and demonstrating strong math/computational background allowed them to work with some pretty top profs doing computational neuroscience. They probably weren't as close to professors as PhD students but definitely close enough to get stellar recs.  Not sure how well this generalizes to other departments.

Edited February 9, 2021 by trynagetby