Much of our attention is on studies of what people know in western democracies. Hell, let's get to the point -- most studies are about political knowledge in America. You'll come across a few UK studies, maybe one from Europe or Australia or even South Korea, but little else.
Along comes this little 2008 gem from the International Journal of Public Opinion Research, which uses Afrobarometer data gathered in Mali to investigate the predictors of political knowledge.
The author, Erik Nisbet, notes some key differences in his findings from those seen in other studies. First off, newspaper reading has no direct effect on political knowledge, unlike in western democracies. This is probably due to lower literacy rates, a hypothesis supported by positive relationships seen between radio and TV news use and knowledge gain. There is an interaction effect for newspaper and education, suggesting that people with higher levels of schooling do get information from newspapers. This makes sense, and I've seen it elsewhere.
What's more interesting to me is the lack of a relationship between any form of media use and "democratic orientations" (supporting democratic ideals). There is also some fascinating discussion of the "knowledge gap" that is the main thrust of the piece.
I also found interesting the results on political participation. For people low in education, listening more and more to radio news did little or nothing to increase their political participation. Higher educated respondents, though, showed a remarkable increase as their radio use increased. Again I've seen similar stuff in my own data, but this suggests that anyone studying media consumption habits and political factors need to look hard at these interactions, which sometimes remain hidden in the data.