One comment I got on my research directions presentation was that my measure of intrinsic motivation to study math was not measuring motivation. I called it “liking math” in the chart I put together, but what is it really? It is composed of these four items:

- I would like to take more math
- I enjoy learning math
- Math is boring (reversed)
- I like math

I should probably call it “positive affect towards math” or better “positive attitudes towards math.” TIMSS calls it the latter (PATM) though they have created a transformed index based on the raw item scores. They report PATM as low, medium, high while I just averaged the scores. Their measure has less information than mine.

You might measure intrinsic motivation with items like this, from the Intrinsic Motivation Inventory:

- I enjoyed doing this activity very much
- This activity was fun to do.
- I thought this was a boring activity. (R)
- This activity did not hold my attention at all. (R)
- I would describe this activity as very interesting.
- I thought this activity was quite enjoyable.
- While I was doing this activity, I was thinking about how much I enjoyed it.

Seems to me that the “positive attitudes towards math” and intrinsic motivation are going to be highly correlated. But maybe you could have positive affect towards math but not much intrinsic motivation? I can’t really see that.

I don’t think this distinction is critical to my project. What I’m investigating is how country-level cultural values influence “returns” to a positive attitude towards math that can be invested in math study as a sort of raw material. I would prefer to look at intrinsic motivation but I’m stuck with what’s in my data set.

Anyway, the issue I’m trying to explore is how different contexts (in this case different cultural values) better support or discourage higher achievement in math for students that have good raw material for learning math.

One potential problem is that positive attitudes towards math are partially caused by prior success in math. So if you’ve been unable to succeed in math in the past, you’re going to have lower PATM. Thus the distributions of PATM across countries could be very different; countries with poor educational systems might have positively skewed (left-leaning) PATM distributions while countries with good ones might have negatively skewed PATM distributions. I’m not sure how this would affect slopes of country-specific math achievement on PATM regressions.