Discussion: View Thread

Hi All,

I'll be teaching a graduate class on statistical consulting this fall semester. I'll be using Cabrera and McDougall's text, along with Hahn and Doganaksoy, and Derr. But I'm looking for any other suggestions for decent readings - something from the idea of "what it's like to be a statistical consultant (or consultant in general)" or "a day in the life of" or even "top skills to have" type readings.

Any suggestions are greatly appreciated.

Thanks,
Michael



-------------------------------------------
Michael Mahometa
Consulting Manager
UT-Austin, Div of Statistics + Scientific Computation
-------------------------------------------

I really enjoy the series that is published in Amstat News. It gives me lots of different ideas of where statistics is being done, and therefore, the types of industries or organizations that might need a consultant.  You might wish to begin with the September 1, 2009 issue which has "the Day in the Life of an Undergraduate Statistical Consultant".

I also believe that a successful applied statistician often works as a vendor to the subject-matter experts. Being able to speak their language, whether the language is physics, biology, social sciences or business, is one of the keys to success.  Giving your students exposure to as many different possibilities now will give them a sense of the pervasiveness of statistical consulting outside academia.

-------------------------------------------
Catherine Trapani
Fordham University, Psychometrics Program
-------------------------------------------

Andrew Gelman of Columbia U was collecting "day in the life of a statistician" writeups that he hoped to use on an ASA blog.  He didn't get enough to use as a continuing feature on the blog, but might have enough for your purposes.

-------------------------------------------
Michael Kruger
Information Resources Inc
-------------------------------------------

One that goes a way back is James R. Boen and Douglas A. Zahn the Human  Side of Statistical Consulting.  JB
-------------------------------------------
John Bartko
Consulting Biostatistician
-------------------------------------------

Hi Michael,

I don't have any suggested readings for you, but I do have a suggestion (and an offer). Rather than read about a statistical consultant, let your students talk to and question a statistical consultant. (And if you need a guest consultant, I volunteer one or more of my students over Google Hangout or Skype.)

In my class, "Communication in Statistical Collaborations" at Virginia Tech, each year during one of my class periods I invite some of my student statistical collaborators from LISA (current grad students who were students in the class the year before) to answer questions from my classroom students. The LISA statistical collaborators that came to my class last year had worked on 30-50 collaboration projects, so they had plenty of experience and stories to draw upon while answering questions. This worked so well last year that I did it twice, brining in a more senior student (who had worked on 80 projects) for a second round of Q&A later in the semester. 

The way it actually worked was that about a week before the event I assigned homework to each student (I had 52 students!) to think of a question they wanted to ask a LISA statistical collaborator. Then in the next class period, teams of 5-6 students discussed their individual questions and picked the one question (or combination of questions) that their team most wanted to know the answer to. I sent this list of 9 most important questions to the guest lecturers a few days before their appearance in class. During the class the LISA collaborators answered the questions and follow-up questions.

The best thing for me was hearing my "lessons" explained to my current students by a former student. Hearing in someone else's words from someone else's mouth really helped the material sink in with the current students.

In summary:

1. Get your students to think about what they really want to know about the daily life of a consultant.
2. Invite a guest lecturer (ideally a former student) to answer questions from current students about statistical collaboration.
3. Current students LOVE the ensuing Q&A and follow-up discussion.
4. I have about 10 current grad students with lots of statistical collaboration experience (40 or more projects worked on) who would jump at the chance to impart their "wisdom" to your students. We could set up a Google Hangout to do this virtually if you don't have enough locals with the right experience.

-------------------------------------------
Eric Vance
LISA (Virginia Tech's Laboratory for Interdisciplinary Statistical Analysis)
Director and Assistant Research Professor
Blacksburg VA, United States
-------------------------------------------