Evaluating Graduate Programs in Statistics
As a professional who uses some basic statistical tools and who is looking at potentially pursuing graduate work in applied statistics (Masters) for use in industry, it would be helpful to know how to best evaluate the various graduate-level statistics programs. What should be the criteria to assess the quality of a university's statistics graduate program? And is there any significant differences in the quality of education received from available online degree options?
2 Replies
If you aren't close to a university that offers a Masters in Statistics, the online option is the way to go.  Online programs can be synchronous or asynchronous. Synchronous programs allow you to attend the class online in real time and ask questions, but it may be difficult to attend a class if you have a business trip or meeting scheduled during class time.  Marshall University and Univ. of South Carolina have synchronous grad Stats programs, while University of Delaware's is asynchronous, so you can watch a video at your convenience.  Some asynchronous schools don't have videos, you simply learn via textbook and communicate with the professor and other classmates through email and message boards.   

As for tuition, schools such as NC State, South Carolina, and Texas A&M charge out-of-state tuition for online programs, while Penn State does not.  There is always the Graduate Certificate option, which typically requires 15 credit hours vs the usual 30 or so for the full degree.  If you live in Arkansas, Georgia, Kentucky, or South Carolina, you may apply to schools in one of the other states through the Southern Regional Education Board's Academic Common Market in order to obtain a tuition credit, which will bring tuition down to the in-state tuition rate.  I hope this helps.
Hi Charles, there are a number of different flavors of programs, so depending on what you are looking for as a career there are different things you might want to look for. More traditional statistics programs (often run out of a math department) may be quite theoretical. Others may be run out of a business school and have a focus on business applications and analytics. Stand-alone statistics departments usually have several types of masters programs available. Many schools have now developed programs more with a data science focus, which is more programming.  Some programs only offer Masters, no PhDs. If you want to stay in the quality/engineering/industrial field, make sure there are classes available and statistics professors with interests in quality/engineering/industrial applications at that department.

A few other key things I might look for as an applied statistician would be the following:
  1. A wide variety of coursework available (online & onsite) -- larger departments will be able to offer more options and be able to provide a more tailored program if you have specific interests, e.g., quality, reliability. If you don't know what you want, having program flexibility lets you figure it out as you go. You should have the ability to adjust if needed.
  2. Design of experiments (DOE) - this may not be a required class any more in many departments but I think it should be mandatory - I personally wouldn't hire anyone without it, especially not into an industrial/engineering environment.
  3. Availability of a consulting office or workshop of some sort, requiring some sort of involvement by the students, or projects. The more involvement you have with solving real problems, the better. It gives you more perspective on what is useful and how to work with people in industry.
I don't have a lot of knowledge about online options specifically, so can't add much to what Geoff posted on that.
Hope this helps and best of luck, sorry for the delayed reply. Feel free to email me directly at chair@asqstatdiv.org if you have any additional questions.

Mindy Hotchkiss
ASQ Statistics Division Chair 2019