The extent of error likely to occur with each of several approximations for the standard deviation, internal consistency reliability, and the standard error of measurement is analyzed. Approximations were compared with exact statistics obtained on 85 different classroom tests constructed and administered by professors in a variety of fields. Means and standard deviations of the resulting differences supported the use of approximations in practical situations. Results of this analysis suggest a greater number of alternative formulas that might be employed, and provide additional information concerning the accuracy of approximations with non-normal distributions. (Author)