On Tuesday night in the University Center, Tom Siegfried, Editor-in-Chief of Science News, delivered the seventeenth annual Alfred and Julia Hill Lecture titled "Odd Are, It's Wrong - The Misuse of Math in Science, Medicine and the Media." Considering the prominence of math in the fields of science and medicine, Siegfried came prepared for journalistic battle.

Siegfried, author of over 900 published articles and three books, focused primarily on the misuse of statistics during his talk.

"The answers that math gives are often misinterpreted.-Tom Seigfried, Editor-in-Chief of Science NewsThe answers that math gives are often misinterpreted," Siegfried said.

Siegfried began by giving the recipe for what makes science "news," (ie. the first report of something, advances in a hot research field, and an issue that contradicts previous belief.) He then gave the recipe for incorrect science, which happened to be the same as what makes science "news." 

Siegfried made the point that journalists write in the process of the scientific discovery they're reporting on and not during the final analysis, unfortunately leaving room for error.

If a sports journalist waited until the end of a basketball season to write their articles, their job would be easy. Predictions would be non-existent. Mistakes would be unacceptable, Siegfried said.

In terms of significance, researchers, journalists and readers sometimes use faulty reasoning. Say the FDA comes out with a new drug, Drug A. According to research, there is a 5 percent chance that Drug A will not work. This does not mean that there's a 95 percent chance that Drug A will work.

"Statistical significance is not the same thing as significance. -Tom Siegfried, Editor-in-Chief of Science NewsStatistical significance is not the same thing as significance," Siegfried said.

Siegfried used a baseball and steroids example to exemplify this faulty reasoning. Out of 400 hundred baseball players, 20 are users (5 percent) and 380 are not users. Of the 20 users, 19 would be identified correctly (95 percent). Of the 380 non-users, 19 would be incorrectly indicated as users (5 percent).

From this surface statistical data, one would assume that the steroid test in use has a 95 percent accuracy rate. Wrong. As Siegfried so poignantly stated, "It's bogus."

Let's look at the math. So, there are 20 users and 380 non-users, as well as 19 indentified users and 19 falsely identified non-users. Thus, we have 38 (19+19) testing positive. Therefore, the probability of any one player testing positive as a user is roughly 50 percent (38/20).

Siegfried suggested that journalists and readers should be more aware of such faulty reasoning. He also advised the audience to remember that there is a chance reports could be wrong.

"We are all guilty [of faulty reasoning]...I am trying to talk people into being less guilty," Siegfried said.