Measuring "Smart"

I suspect a decision by the Armenian government to introduce chess as a school subject for primary school students will begin the debate about what it might achieve. The general consensus from within the chess community is that chess makes us smarter. However what we think, and what non-chessplayers think, may not carry much weight without some objective measurements attached.

I have a particular interest in the decision by Armenia, because chess is already a subject for grades 3,4 &5 at my childrens school, as well as being an elective subject for 6th grade students. Each class gets an hour a week of chess instruction, and the program has been running for a little over a year. While it is a popular program, and anecdotal evidence points to the students improving their problem solving skills and concentration levels in their other classes, the difficulty is measuring this improvement.

Some ideas include comparing exam marks for classes doing chess against classes prior to the programs introduction, although discounting other factors may be difficult. Studying second order effects such as behaviour and attendance may also be possible (eg see if attendance levels are higher on days when chess classes are held).

In fact it may may a good academic study to not only see it chess makes us smarter, but what metrics are used to measure this.