'G-rated' talk to address impact of Simpson's paradox on everyday life


11 a.m., Jan. 28, LL2

Are statistical analyses that affect major decisions made in our society the real picture? Whether it is medical studies, mental health services, batting averages or unemployment rates, individuals often are affected by the research results of social and medical studies. Yet often many of these studies produce inaccurate conclusions based on faulty analysis of the data collected, according to Harvard professor Xiao-Li Meng.

Few paradoxes have impacted everyday life more than Simpson’s Paradox, according to Meng, yet, paradoxically, Simpson’s paradox is not even a paradox in the mathematical sense.

Meng, the Whipple V. N. Jones Professor of Statistics and chair of the Department of Statistics at Harvard, will present a seminar and discussion on the danger of Simpson’s Paradox in his talk “Trivial Mathematics but Deep Statistics: Simpson’s Paradox and Its Impact on Your Life.” The program will begin at 11 a.m., Jan. 28, in the Languages and Literature building, room 2 (LL2).

Having a statistical background is not necessary to understand this talk, though common sense and a desire to look beyond the formulas is recommended. This is also a G-rated talk – a “gadgeted” seminar. Never heard of it? This is your chance, says Meng.

This statistical seminar series is sponsored by the School of Mathematical and Statistical Sciences and the Mathematical, Computational and Modeling Sciences Center in ASU’s College of Liberal Arts and Sciences. It is free and open to the public. An abstract and more information on Meng and his research are online at http://math.asu.edu/node/3437. A reception to welcome Meng to ASU will be held at 10:15 a.m. in the Physical Science Center A-wing, Room 206.

For more information on this and other statistics seminars presented by the School of Mathematical and Statistical Sciences visit http://math.asu.edu/events or contact 480-965-9792.

Caleen Canady, cali@asu.edu
480-965-9792
School of Mathematical and Statistical Sciences
College of Liberal Arts and Sciences