Theoretical vs applied statistics

Anish Thomas writes:

I was wondering if you could provide me with some guidance regarding statistical training. My background is in Industrial/Organizational Psychology, with an emphasis on Quantitative Psychology and currently working in the employee selection industry. I am considering pursuing a masters degree in Statistics. As l look through several program options, I am curious about the real difference between theoretical and applied Statistics. It would be very enlightening if you could shed some light on the difference. Specifically:

1. Is theoretical side more mathematically oriented (i.e., theorems and proofs) than applied?

2. Are the skills acquired in a ‘theoretical’ class difficult to transfer to the ‘applied’ side and vice versa?

3. I see theoretical statistics as the part that engages in developing the methods and applied statistics as pure application of the methods. Is this perception completely off base?

My reply:

1. The difference between theoretical and applied statistics is given here.

2. I think the skills are highly transferable, except at the extremes. The strong law of large numbers won’t help you with applied statistics, and telephone interviewing techniques won’t help you with probability theory. But just about anything in between, from stochastic processes through statistical inference through the design of survey weights, should be good for anybody with theoretical or applied interest.

3. No. Applied statistics includes development of methods as well.

2 thoughts on “Theoretical vs applied statistics

  1. At Indiana University, the only substantive difference between the MS in Applied Statistics and the MS in Statistical Science is that the latter requires a (Masters) thesis, while the former does not.

Comments are closed.