Ergodicity is a concept created at the end of the 19th century, which defines a kind of stochastic process and is very useful in many applied sciences including human sciences. Its main premise says that population and individual averages are equal. In psychometrics, it has been common to use population average to make inferences about the measure of the individual. Indeed, this can be done, but exclusively in ergodic processes and systems. However, none of the processes involving human beings, be they mental or physical, can ever be ergodic. This problem was first highlighted by Molenaar, but later, many researchers including Molenaar began publishing about processes that could be interpreted as not fully ergodic; the literature calls this a “continuum,” as opposed to “all or nothing” which we argue is the only correct approach. We claim that a process is either ergodic or nonergodic, there is nothing in between, and that time is ripe for researchers working in all sciences related to human behavior and health to acknowledge this truth: population measures cannot be used as measures of the individual.
Gomes et al. (Tue,) studied this question.