Posts written by Chris Wild (2)

avatar

Chris Wild did his PhD at the University of Waterloo in Canada before joining The University of Auckland 1979. He is a Fellow of the American Statistical Association, a Fellow of the Royal Society of New Zealand, and an Editor of the International Statistical Review. He has been a Council member of the International Statistical Institute, President of the International Association for Statistics Education, an Associate Editor of Biometrics, the Statistics Education Research Journal (SERJ), and the Australian and New Zealand Journal of Statistics. He was Head of Auckland's Department of Statistics 2003-2007 and co-led the University of Auckland's first-year statistics teaching team to a national Tertiary Teaching Excellence Award in 2003.

His main research interests are in developing methods for modelling response-selective data (e.g. case-control studies) and missing data problems, and in statistics education with particular emphasis on statistical thinking and reasoning processes.

June 17, 2011

“Shocking world of our student drunks” – Where did that come from?

“Shocking world of our student drunks” shouted the 10 June front page headline from the New Zealand Herald.

“Nearly a third of university drinkers have passed out while boozing in the past six months”, it continued. Moreover, “27 percent of men and 9 percent of women say throwing up will not stop their boozing.”

And where did all that come from?

The figures came from a survey of students at just three student dormitories at a single university, and could therefore be symptomatic of a very localised culture, but were presented as picture of student behaviour across an entire country.

The authors of the source paper in the New Zealand Medical Journal carefully reported that their research population was “three student residential facilities in 2006”. In addition to their place of residence these students were very unrepresentative of the student body as a whole in terms of age and sex.

Most percentages quoted were related only to drinkers but sound in the news report as if they were percentages of all students.

Also 40% of those contacted did not take part. Ignoring nonresponse biases the male figures have an unacknowledged margin of error of the order of 7%.

So what is our point?

An implied applicability of the results that goes far beyond what is justified from the research undertaken and oversensationalising to make an attention grabbing story.

April 7, 2011

This cartoon is good

“So, uh, we did the green study again and got no link. It was probably a– RESEARCH CONFLICTED ON GREEN JELLY BEAN/ACNE LINK; MORE STUDY RECOMMENDED!”