Skip to main content
. 2017 Jan;146(1):63–76. doi: 10.1037/xge0000245

Figure 13.

Figure 13

Each data point is 1 of 16 possible subsequences of length 4. The rate with which participants generate each of these sequences is poorly predicted by entropy, but well predicted by the Hahn and Warren (2009) account. The postexposure generated occurrence rate fit with Hahn and Warren is r = .92 versus 0.67 for Entropy.