random effect
an effect arising from an independent variable in an experimental design whose values or levels are drawn randomly from some larger (conceptual) population of levels that could (in principle) have been selected. For example, a health researcher investigating the relationship between exercise and weight may select a few levels of daily exercise for study (e.g., 0 hours, between 0 and 1 hour, between 1 and 2 hours, between 2 and 3 hours) from a wide range of possible options. Results involving a random effect can be generalized to values beyond those observed and modeled in the study analysis. Also called random factor. Compare fixed effect.