I decided to test the relationship between emotionality and wealth as measured by PPP-adjusted GDP per capita, predominant language (a proxy for colonial influence by France or England), and whether at least half the population was Muslim. I used a linear OLS regression; results are below:
People in Francophone Africa were, on average, 3 percentage points less likely to report expressing or feeling emotion, a result that was statistically significant. This may not sound like much, but it's a whole standard deviation below the mean emotionality score. (The model also explains more than a quarter of variation in African countries' emotionality scores!)
So why are the Francophones less effusive? A couple hypotheses:
1. Maybe when translated into French, the survey was less likely to result in positive answers for some reason.
2. The Francophone variable might actually be capturing regional variation, such as lower emotionality in West Africa.
3. The French colonial cultural influence may have included elements of less emotional expression.
The first issue is hard to address (language and cultural translation is always an issue when trying to compare social tendencies across cultures.) Putting in regional variables could address the second issue. Another interesting thing to look at would be the impact of political instability. My data, in .dta format for stata, are here: /uploads/1/7/1/1/1711915/africa_emotionality.dta