Experimental design and analysis illustrated with methadone. Drugs were tested in separate groups of 12 mice (6 per sex) using a within‐subjects repeated‐measures design. The left panel shows that initial analysis pooled raw data from both sexes. Abscissa: dose methadone in mg/kg administered SC. Veh = vehicle. Ordinate: Total locomotor activity counts during a 60‐min test session. Bars show mean ± SEM, and points show individual data. **indicate different from vehicle as indicated by one‐way ANOVA followed by a Holm‐Sidak post hoc test, p < .01. The middle panel shows calculation of dose‐effect parameters (E
max, ED50). Data for each mouse at each dose were transformed to % Methadone E
max using the equation [(Drug − Veh)/(5149.8)] × 100, where Drug = total locomotor counts in a given mouse after a drug dose, Veh = mean locomotor counts after vehicle in that group, and 5149.8 = the mean maximum increase in locomotor counts produced by methadone in the methadone group (7500.1 counts at 32 mg/kg vs. 2350.3 counts after saline vehicle). Filled symbols indicate different from vehicle as indicated by one‐way ANOVA followed by a Holm‐Sidak post hoc test, p < .05. The right panel shows the same data as the middle panel segregated by sex and analyzed by two‐way ANOVA. In this case, there was a main effect of dose [F (2.82, 28.2) = 38.74, p < .0001], but no main effect of sex [F (1, 10) = 0.42, p = .53] and no sex × dose interaction [F (4, 40) = 0.98, p = .43].