Results of my non-scientific estimation experiment are in. Participants were randomly divided in two groups. One group was asked two questions:
A1) Are there more than 10 countries in Africa?
A2) How many countries are there in Africa?
The other group was asked two questions:
B1) Are there more than 100 countries in Africa?
B2) How many countries are there in Africa?
The average # of countries guessed by group A was 25.6, standard deviation 16.7 (20 total responses).
The average # of countries guessed by group B was 55.9, standard deviation of 22.8 (18 total responses).
There is less than a 1% chance that the two groups of responses are similar (according to a T-Test, if you're statistically inclined). 50% of the responders' answers fall within the boxes in the figure.
Sadly I didn't make up this experiment; I just wanted to see if I could duplicate the results. This psychological technique is called "priming", or sometimes "anchoring". It's the same concept at play when you see a price tag with the "suggested price" crossed out and the reduced price shown below. People are so easy to manipulate!