Figure 5.

Population-level similarity (a) and effects of population size (b) in Model 3. (a) Similarity between items and their inputs. ‘Biased sampling’ with k = 0.1 and ‘Convergent transformation’ as in Study 1. Results are averaged on 10 runs of the model. The shaded area shows standard deviations. In both conditions N = 100. (b) Time to reach equilibrium for different population sizes. Measured for ‘Biased sampling’ at two different levels of k (k = 0.1; 0.5) and for ‘Convergent transformation’. Results are averaged on 10 runs of the model. Bars show standard deviations.