fixed effect
an independent variable whose levels are specified by the researcher rather than randomly chosen within some level of permissible values. For example, a health researcher who specifically chose to examine the effect on weight loss of no exercise, 1 hour of exercise, or 3 hours of exercise per week would be treating time spent exercising as a fixed effect. In other words, all levels of interest are included in the design, and thus anyone wanting to replicate the study would have to use the same levels of exercise as in the original. Also called fixed factor. See fixed-effects model; random-effects model. Compare random effect.