You’re probably disappointed to hear that this post is not about my drunken exploits in my first month as a PhD student. It’s about science, specifically a recent Stanford study on first year students’ attitudes toward alcohol.

The researchers asked the students about their views on alcohol use, how they thought their views compared to other students’ views, and their willingness to discuss their views with others. The authors found that:

…college students who were extremely pro-alcohol were more likely to express their opinions than others, even though most students surveyed were moderate in their views about alcohol use.

They also offer a description of the kind of environment that encourages these opinionated extremists to come out of the woodwork:

How do people with extreme views believe they are in the majority? This can happen in groups that tend to lean moderately in one direction on an issue. Those that take the extreme version of their group’s viewpoint may believe that they actually represent the true views of their group

This finding chimes nicely with the ‘group polarisation’ tendency discussed in a working paper by Sunstein & Hastie. Unsurprisingly, speculation about real-world occurrences abounds: the original news source suggests it could explain extreme opinions in US politics, whereas the I09 blog sees it as a good behavioural model of internet trolling. Any more ideas?