another LOR from a professor or are PhDs outside academia all right?

[clurp]

I'm applying for stats PhDs and I'm having some last minute second thoughts about how I've chosen my 3 recommenders. 

 

A little background: 

 

• I did math as an undergraduate (I had a > 3.7 GPA at a 'top 20' university, with math grades ranging from B to A+)  
• I took a pretty tough schedule with a lot of upper division and graduate courses, and I participated in two summer research projects. 
• I studied for an applied math masters in Europe for a year.
• During that year, I took a sequence of courses there that got me interested in statistics.
• The summer after the first year there, I got a job at a startup doing statistics and machine learning.
• I liked the job a lot and knew that I ultimately wanted to do a stat PhD, so I stayed beyond the summer and did not finish the masters.

For letters of recommendation, I asked one professor that I knew quite well from coursework and summer research at my undergrad school, and I asked two PhDs (one in math, one in statistics, both from very good schools) that I work with at my current job. One has published in reputable math and stats journals and was a university instructor before his career doing quantitative research in industry. The other has published some as well, but I know less about that. The idea was that I wanted to ask the people who know me well, that I wanted to try to highlight the work I had done recently  rather than 3 or more years ago, and that it's ok to have non-academic recommenders if they have PhDs and work in research.

 

Am I wrong on that last point? 

 

I'm considering whether I ought to send a request for a letter of recommendation from another undergraduate instructor who I've previously asked for a recommendation. I got an A+ in an upper division math course of hers, although it wasn't very closely related to statistics. Should I ask her if she still has a copy of that recommendation on her computer? Is it important to have at least 2 recommendations from people currently in academia? 

 

Of course, it would be best to get another recommendation from the professor of the statistics courses which led me down the path I'm on or the professor of a graduate analysis course, but because I'm not close with those professors, I'd expect the recommendation to be generic.  

Edited November 12, 2013 by clurp

[cyberwulf]

This is fine. Letter writers should have the ability to put your performance into a relevant context and the credibility to vouch for your ability to succeed in a good graduate program. Since it sounds like your workplace letter writers can do both, you don't need to worry.