What are the benefits of getting a PhD in statistics?

Question

I've been working as a statistician in the marketing world for two years, and I have an M.S. in applied math. I want to change industries a bit (maybe epidemiology or engineering) and I also want to live overseas (ideally, a job where I'd live in one country for a few years and then move on to another).

So, my question is: to accomplish these life goals, does it make sense to get a PhD? Or, is there some other route that would be more beneficial?

Obvious question but...have you considered a PhD in Epidemiology? — Fomite, May 23 '16 at 20:04

score 13 · Accepted Answer · answered Jul 14 '13 at 17:16

13

A doctoral degree is a credential signifying largely that you are capable of doing independent research at the highest possible level. It is not, in an of itself, a ticket to working in a particular industry or in a particular location any more than a bachelor's or a master's degree. In fact, PhD holders may have substantially more challenges in those aspects, because the additional qualifications make them unattractive for many positions in conventional businesses and industries. (You are unlikely to find a PhD working in a sales division of a multinational conglomerate, for instance.)

The reason to get a PhD is because you are interested in problem solving and doing original work. If this doesn't describe your motivation, I would recommend against pursuing a graduate degree, because it will be a very long few years of your life which are not guaranteed to achieve the objectives you've laid out.

answered Jul 14 '13 at 17:16

aeismail

173,481
34
418
736

1

So, if I understand correctly, the PhD is only valuable as a stepping stone to a career in research/academia? It doesn't really have much value for industry purposes? – random_forest_fanatic Jul 15 '13 at 01:34
8

In most industries, a PhD has negative value. – JeffE Jul 15 '13 at 02:27
3

@JeffE: Outside of research divisions. – aeismail Jul 15 '13 at 04:19
1

@JoshBrowning: There are other career paths also: scientific project management, but also things like patent law, scientific journalism, and start-ups. – aeismail Jul 15 '13 at 04:22
3

@JoshBrowning: I'd say a PhD is a stepping stone to a career in research or research-related work. That may also be outside academia. – silvado Jul 15 '13 at 07:07
Thank you all for all the helpful advice. It sounds like the question I have to answer now is "Do I want to do significant research in my future career?" – random_forest_fanatic Jul 15 '13 at 11:28
@JeffE do you have a reference/source for this claim? – Jase Aug 20 '13 at 10:13

score 6 · Answer 2 · edited Apr 13 '17 at 12:44

A PhD in statistics is more flexible and useful that PhDs in some other areas. The usual issue with PhDs one hears about is that one becomes over-qualified for non-academic work once one has a PhD. Additionally, there is a lot of time spent getting it.

However, statistics is intrinsically an applied science, and one that is in big demand across lots of areas, because it can be applied to lots of areas, unlike most academic disciplines. Specific anecdote: I was once told by a Statistics Professor that the head of a clinical trial is required to have a PhD in statistics (by the NHS, possibly). I don't know if this is true, but it sounds like something that is probably true. As he put it, this creates jobs for PhDs.

With computers being used more and more, and lots of data being created that needs to be analysed, new methods need to be invented to handle all this data. This is the kind of quasi-research work which is quite well suited for someone with a PhD.

Areas like data visualization and graphics are quite hot right now. Having a PhD in an area like that will probably not hurt you. See Hadley Wickham's thesis for example.

Of course, it is possible to get a PhD from a Statistics Department without learning any statistics, for example if you write a Probability (Mathematics) thesis. You probably don't want to do that.

My personal experience (I have a Statistics PhD) is that to get an interesting work, even in industry, a PhD is helpful. Much of the work so-called statisticians do is to mindlessly apply standard algorithms from some software package to data using things like SAS, and then package up the (machine produced) results. If you have a functioning brain, you don't want to do that.

BTW, it seems such questions are not on topic at stats.sx, but you could ask people on chat there - perhaps point to this question.

Thanks Faheem, I'm glad to hear from another statistician! I'm considering focusing my PhD on spatial statistics, as I'm interested in epidemiology and geographical applications. Do you have any opinions on that particular topic? — random_forest_fanatic, Jul 20 '13 at 11:43
Hi Josh. No, I don't know anything about spatial statistics. Have you thought about which dept you want to go to? Choosing a dept carefully is important. As is the adviser, of course. You should also consider, of course, what your aim is in getting the PhD. As I said, getting a statistics PhD does not necessarily constrain you to academia, but what you work on in your thesis should at least somewhat depend on what you want to do afterwards. — Faheem Mitha, Jul 20 '13 at 15:24

What are the benefits of getting a PhD in statistics?

2 Answers2

Linked