This paper expands upon a previously reported random speckle technique for measuring the modulation transfer function of midwave infrared focal plane arrays by considering a number of factors that impact the accuracy of the estimated modulation transfer function. These factors arise from assumptions in the theoretical derivation and bias in the estimation procedure. Each factor is examined and guidelines are determined to maintain accuracy within 2% of the true value. The uncertainty of the measurement is found by applying a one-factor ANOVA analysis and confidence intervals are established for the results. The small magnitude of the confidence intervals indicates a very robust technique capable of distinguishing differences in modulation transfer function among focal plane arrays on the order of a few percent. This analysis directly indicates the high quality of the random speckle modulation transfer function measurement technique. The methodology is applied to a focal plane array and results are presented that emphasize the need for generating independent random speckle realizations to accurately assess measured values.