ON SAMPLE-SIZE, STATISTICAL POWER, AND THE DETECTION OF DENSITY DEPENDENCE

In reviewing the results of Stiling (1988; see 89L/02762) on the detection of density dependence in insect populations, Hassell, Latto & May (1989; see 90L/03758) observed that density dependence is more commonly detected in large samples. This observation is here interpreted in terms of the statistical power of the test used to detect density dependence. Using simple models, the authors show that even for moderately strong density dependence, it is necessary to observe up to 30 generations before density independence can be rejected with high probability. -from Authors