Collaborative Study of a Fluorometric Method for Determining Selenium in Foods
Abstract A fluorometric method for the determination of selenium in foods reported previously was studied collaboratively. Nineteen laboratories reported analytical results on 10 samples representative of vegetables, cereal, dairy products, meat, and fish, and containing naturally occurring selenium. For the 5 pairs of samples, the coefficients of variation based on precision standard deviation were 2 5 . 9% at 16 ng selenium, 15.8% at 98 ng, 8.5% at 233 ng, 6.1% at 379 ng, and 4.1 % at 427 ng. The corresponding coefficients of variation computed from overall standard deviations were 65.3, 17.5, 8.9, 8.1, and 5.6%. Systematic error was significant for 2 pairs (F-test, p = 0.01), but not for the remaining 3 (p = 0.05). The ratio of the systematic error standard deviation to the precision standard deviation varied from 0.21 to 1.64. Analyses of 2 NBS standard reference materials, orchard leaves (75 ng selenium) and bovine liver (208 ng selenium), yielded results higher by 16 and 10%, respectively. The precision and accuracy of the collaborative procedure were deemed acceptable and the method has been adopted as official first action.