Cytel has taken the initiative to train the next generation of clinical programmers through its innovative Clinnical Programming Laboratory [see Cytel's CliPLab].
What about training the next generation of statisticians? The Harvard School of Public Health has just awarded the 2014 Marvin Zelen Leadership Award in Statistical Sciences to a distinguished statistician and educator.
Last Friday, Xiao-Li Meng accepted the 2014 Mavin Zelen Leadership Award in Statistical Sciences. Meng is the Whipple V.N. Jones Proefssor of Statistics at Harvard University, and currently serves as the Dean of Harvard’s Graduate School of Arts and Sciences. He joins an illustrious list of past Zelen Award recipients including Frederick Mosteller, John Tukey and Sir David Roxbee Cox.
The Marvin Zelen Leadership Award is named for distinguished statistician and Cytel Board Member, Professor Marvin Zelen, whose many contributions to public health include conventionalizing clinical trial methodology in the 1960s to take randomization, trial design and data quality into account. Zelen was the first chair of Harvard’s Department of Biostatistics, and served as the Head of the Division of Biostatistics at the Dana-Farber Cancer Institute. As many Cytel blog readers will know, he also posited the statistical problem that led to the foundation of Cytel Inc. [See 'A Horizon for the Stars'.]
Like Marvin Zelen, Xiao-Li Meng is not only an eminent statistician but also a passionate educator. He teaches on the first year doctoral course, ‘The Art and Practice of Teaching Statistics,’ and was the lead developer of the undergraduate course, ‘Real-Life Statistics: Your Chance for Happiness (or Misery).’ The aim of this course was to introduce students to statistical reasoning without appealing to abstractions and technicalities.
In honor of this teaching style, here is a statistical puzzle to which Professor Meng would refer in his introductory course, during the first lecture of the year:
Which Treatment is Preferred?
Suppose you are a doctor whose patient is suffering from kidney stones. You have the choice of two treatments, Treatment A and Treatment B. Both treatments have been tested on 350 patients.
Treatment A was tested on 87 patients with small stones, and was successful for 81 of them. It was also tested on 263 patients with large stones and was successful for 192 of them.
Treatment B was tested on 270 patients with small stones and was successful for 234; it was also tested on 80 patients with large stones and was succesful for 55.
As a result we know that: Treatment A is successful 78% of the time and Treatment B is successful 82% of the time. We also know that for patients with small stones, Treatment A is more successful than Treatment B (93% and 87% respectively.) When patients have large stones, Treatment A is also more successful than Treatment B (73% and 69% respectively).
Suppose you do not know whether a patient has large stones or small stones. Which treatment is preferred treatment?
Related Items of Interest
Xiao-Li Meng, Statistics: Your Chance for Happiness (or Misery)
Marvin Zelen, The Training of Biostatistical Scientists
