Does there exist a macroeconomic model in the head of a microeconomic agent?

I have a simple test.  Select 1,000 adults from a random selection of middle class home.  Have them take a survey with the following question:

Describe a list of roads that automobiles may travel on.

Look at the subsequent lists generated.  Does the list generally distinguish between local streets and interstate highways.  If we find that lists distinguish between local roads for short trips and highways for cross country long trips, then we can say the list partitions the aggregate transportation system in macroeconomic model.  We have a spectral partition, we have a calculus, we have a norm, a fixed equilibrium point,  bounded functions and we have a macroeconomic model, however imprecise, in the agents head. 

If we talk about agents finding patterns for trade or work, then I think we are stuck with agents finding macroeconomic patterns.