Q-test. Use to identify statistical outliers in data. This test should be applied sparingly and never more than once to a single data set.
R is the range of all data points, xa is the suspected outlier, and xb is the data point closest to xa. At the 90% confidence interval Q for N replicate measurements is:
