Collaborative Study of Determination of Total Nitrogen in Brown Rice by the Combustion Method According to the Dumas Principle

Abstract

The total nitrogen contents of brown rice were measured by the combustion method in cooperation with 16 laboratories in Japan. According to the protocol that determined the measuring conditions etc., six samples of brown rice were measured with 3 repetitions and 3 times on different days. The data were removed with Cochran and Grubbs outlier test, and analysis of variance (ANOVA) were carried out for samples No.3∼No.6, samples No.1 and No.2 were evaluated as the Youden Pairs. The range of analysis values for six samples calculated from a collaborative trial were as follows: Sr (the standard deviation of laboratory repeatability), 0.003∼0.009 g; SR (the standard deviation of laboratory reproducibility), 0.019∼0.031 g; RSDR (relative standard deviation of reproducibility), 1.73∼2.56%; and HorRat values, 0.44∼0.66, respectively. Nested random effects analysis of variance for values of samples No.3∼No.6 was conducted and the total variation of ANOVA was divided into three sources of error; 85∼93% of the SR contributed to an inter-laboratory error, and contribution of the intra-laboratory error between different days (3∼8%) and error with in a sameday (2∼7%) were very small. HorRat values of all sample being lower than 1.0 show that the measuring results of this collaborative trial were satisfying. Moreover, RSDR of this trial show that the validity of the measuring method of the total nitrogen content in brown rice by the combustion method was certified.