Much of the variation in how people respond to "Humans evolved from earlier species" seems to be due to whether their particular culture/religion requires them to take a specific view on the subject [1]. If it does, people answer in line with their culture/religion, rather than science.
In effect, questionnaires like this end up studying what particular religions have to say on the matter, and it just happens that the US cultures/religions are more backwards (less accepting of science) than others.
The fit of A(1-B/x) through the scatterplot is kind of misleading too. Apart from Turkey and USA, the whole plot is more like 2 or 3 clusters, which are already highlighted, and the actual relationship with GDP per capita is much weaker than implied by the fit curve. Look at the actual points: the colours (country location) contain most of the information, and the fit curve doesn't add much information.
[1] http://www.culturalcognition.net/blog/