Drink that cheers adds to the years


This descriptive article raises the issue of what "signifiance" means and could be used with older students who are starting to perform hypothesis tests and look for "significant differences". Several issues arise but are not answered by the information given.

A study of 3000 people might have made large numbers of people involved in the various groups so that a very small actual difference could be "statistically significant" at say the 5%-level. One should always be suspicious of such differences if enough information is not provided to be able to decide if the difference in life expectancy is great enough to matter in people's life planning.

Another issue is that of cause-and-effect. It appears that the researchers are confident in claiming that it is the alcohol which increases the life span. It would be interesting to know the other variables used as controls in the study so that their potential effect could be eliminated.


Where to next?

Student Questions for this article
Newspaper article
Index - Related articles
Index - Inference
Numeracy in the News - Main Index