What a Statistical Consultant Does
Statistical consultants are professionals who provide advice and guidance on the statistical aspects of scientific and medical studies. This seems obvious, except that
it is not always obvious what aspects of a study will affect the statistics. It is the responsibility of the statistician to know what aspects are relevant and to make
the client aware of them, if necessary. This often depends on the environment in which the study is conducted and who will be reviewing the results.
We also provide services with more tangible results, such as designing experiments, calculating sample sizes, formulating and testing statistical models, estimating
parameters, and creating data listings, tabulations, data analyses, and graphical displays. We try to design studies
that are practical, so that they will be successfully
And we deal with the consequences of those that didn't go as planned.
That is, we give advice and guidance, and we can get our hands dirty as well.
Statisticians turn data into information, and we devise ways to make large, cumbersome amounts of data accessible and understandable.
I hope to become a full member of your team, assisting in every way that I can to develop the scientific information that you need for your product or process. I have no
financial interests in any healthcare or pharmaceutical company other than mutual funds. That way, I can maintain my independence. What you get when you hire me is my
unbiased, independent, professional opinion, based on my training and experience.
I've been around for a while. I have a lot of education in mathematics and statistics. I also have long experience at NIH and FDA. Most of this will remain invisible to
you, but it is still working for you. I am also accumulating practical experience in the pharmaceutical industry, and you will get the benefit of my experience with
clients who preceded you. (Don't worry: I won't give away your secrets or theirs.)
Most nonstatistical scientists know something about statistical practice. Some of them know a great deal. I respect this knowledge because I know that it was difficult to
get. But I believe that you will
be hiring me because my expertise is unavailable to you otherwise.
I can carry out the mechanics of producing statistics (and I do enjoy
this part of the job), in the same way that the nonstatisticians
can. But you will be hiring me for more than that--for my knowledge of what and when and how and why to
produce these particular statistics. This is my unique contribution. It is the value I add that others cannot.
I will need to discuss statistics as well as the scientific area of application with the other professionals on the project. Each of us should understand what everyone is
trying to do. This is particularly important for statisticians, because our efforts have to be defined in the context of the application, which might involve scientific,
regulatory, legal or other considerations.
One thing a consultant doesn't do and shouldn't do is to take over the implementation of the project. That is, I can make a recommendation or design a study or analyze
the data, but it is always the client's decision and responsibility to take action on the recommendation,
or to send the report for publication. Naturally, I will stand
behind my work, but I cannot take the responsibility that rightfully belongs with my client.