Interlaboratory comparability, bias, and precision for four laboratories measuring analytes in wet deposition, October 1983-December 1984

Four laboratories involved in the routine analysis of wet-deposition samples participated in an interlaboratory comparison program managed by the U.S. Geological Survey. The four participants were: Illinois State Water Survey central analytical laboratory in Champaign, Illinois; U.S. Geological Survey national water-quality laboratories in Atlanta, Georgia, and Denver, Colorado; and Inland Waters Directorate national water-quality laboratory in Burlington, Ontario, Canada. Analyses of interlaboratory samples performed by the four laboratories from October 1983 through December 1984 were compared.

Participating laboratories analyzed three types of interlaboratory samples--natural wet deposition, simulated wet deposition, and deionized water--for pH and specific conductance, and for dissolved calcium, magnesium, sodium, sodium, potassium, chloride, sulfate, nitrate, ammonium, and orthophosphate. Natural wet-deposition samples were aliquots of actual wet-deposition samples. Analyses of these samples by the four laboratories were compared using analysis of variance. Test results indicated that pH, calcium, nitrate, and ammonium results were not directly comparable among the four laboratories. Statistically significant differences between laboratory results probably only were meaningful for analyses of dissolved calcium. Simulated wet-deposition samples with known analyte concentrations were used to test each laboratory for analyte bias. Laboratory analyses of calcium, magnesium, sodium, potassium, chloride, sulfate, and nitrate were not significantly different from the known concentrations of these analytes when tested using analysis of variance. Deionized-water samples were used to test each laboratory for reporting of false positive values. The Illinois State Water Survey Laboratory reported the smallest percentage of false positive values for most analytes. Analyte precision was estimated for each laboratory from results of replicate measurements. In general, the Illinois State Water Survey laboratory achieved the greatest precision, whereas the U.S. Geological Survey laboratories achieved the least precision.