Michael Kosorok
Kosorok
A short time ago, I gave a presentation on clinical trial design, precision medicine, dynamic treatment regimes, and sequential multiple assignment randomized trials (SMARTs), to an audience of clinicians. I focused my talk on the ideas and avoided statistical formulas (which can be difficult for me to do). I felt that the audience was attentive and interested. And, based on my conversations with several of them afterwards, they seemed to understand the ideas I was trying to convey.
Around the same time, one of my collaborators received a summary statement for a study we had proposed and for which I had helped to write the statistical aspects. The review included some criticisms of the statistical aspects. While none of us enjoy criticism, what frustrated me most was that the criticisms clearly showed that the reviewer did not understand certain statistical concepts. Moreover, when I looked at the study section roster, there were no statisticians listed!
These two contrasting examples illustrate different aspects of the importance and challenge of communicating about statistical concepts in biomedical research. The first example shows how we can reach across different disciplines to build understanding. Whereas, the second shows that even if we do a great job communicating to one audience, the same approach will not necessarily work with another audience. The second example also suggests that part of the solution may require improving basic training in statistics for all biomedical researchers.
Through these and many other experiences, I have concluded that there are several aspects of statistical communication that are crucial for the success of biomedical research:
- statistical inference is fundamentally important in essentially all areas of biomedical research, and the success of the human race depends on it;
- we as statisticians need to work harder at communicating with non-statisticians to improve collaboration and to increase the chances of scientific success;
- more of us need to participate in scientific review processes;
- there are not enough statisticians with the needed statistical expertise combined with the needed communication skills in the world; and
- there are not enough non-statistical biomedical scientists who understand enough statistics to consistently make good judgements about inferential aspects of research nor to know when they need expert statistical help.
To solve these challenges, we must work harder on communication as a discipline, train more statisticians with the right skill sets, and work harder in educating non-statisticians. It is essential that we seek the help of our colleagues in non-statistical biomedical disciplines to make the needed changes in education for all scientists. And finally, we need to find ways to increase awareness about statistical inference for everyone.
Comments/Discussion
| Submit a question or comment. |
Each month, one of our program investigators will introduce him/herself and will discuss their research, new research directions, or advancements made toward our goal of improving the clinical trial process. Readers are encouraged to send questions or comments. In addition, we will announce new software releases, publications, and upcoming events.
| Subscribe |
Date: February 15, 2016
Date: January 15, 2016
Date: September 17, 2015
Date: July 16, 2015
Date: June 4, 2015