Literature on optical and infrared microvariability in active galactic nuclei (AGNs) reflects a diversity of statistical tests and strategies to detect tiny variations in the light curves of these sources. Comparison between the results obtained using different methodologies is difficult, and the pros and cons of each statistical method are often badly understood or even ignored. Even worse, improperly tested methodologies are becoming more and more common, and biased results may be misleading with regard to the origin of the AGN microvariability. This paper intends to point future research on AGN microvariability toward the use of powerful and well-tested statistical methodologies, providing a reference for choosing the best strategy to obtain unbiased results. Light curves monitoring has been simulated for quasars and for reference and comparison stars. Changes for the quasar light curves include both Gaussian fluctuations and linear variations. Simulated light curves have been analyzed using χ2 tests, F tests for variances, one-way analyses of variance and C-statistics. Statistical Type I and Type II errors, which indicate the robustness and the power of the tests, have been obtained in each case. One-way analyses of variance and χ2 prove to be powerful and robust estimators for microvariations, while the C-statistic is not a reliable methodology and its use should be avoided.