Violations of betweenness or random errors?

A betweenness axiom states that if A and B are equally good then a mixture of A and B is equally good as well. This note demonstrates that the violations of the betweenness axiom documented in several experimental studies can be alternatively attributed to the effect of random errors.