Effects of Sample Size and Sample Acceptance Level on the Number of Aflatoxin-Contaminated Farmers’ Stock Lots Accepted and Rejected at the Buying Point

Abstract

Abstract Sixteen different aflatoxin sampling plan designs were evaluated using the negative-binomial distribution. Evaluations were used to predict the effects of 4 different sample sizes and 4 different sample acceptance levels on the classification of farmers’ stock lots according to the lot aflatoxin concentration. The 4 sample sizes evaluated were 2.27 kg (5 lb), 4.54 kg (10 lb), 9.08 kg (20 lb), and 18.16 kg (40 lb). The 4 sample acceptance levels evaluated were 5,50,100, and 180 ng/g. A decrease in the sample acceptance level from 180 to 5 ng/g decreased the number of lots accepted at all lot concentrations, increased the number of lots rejected at all lot concentrations, and decreased the average aflatoxin concentrations among all lots accepted. At the highest sample acceptance level, 180 ng/g, increase in sample size from 2.27 to 18.16 kg decreased the percent of lots accepted at concentrations above the sample acceptance level and increased the percent of lots accepted at concentrations below the sample acceptance level. At the lowest sample acceptance level investigated, 5 ng/g, an increase in sample size from 2.27 to 18.16 kg decreased the percent of lots accepted at all concentrations. The effect of using sampling designs with 2 sample acceptance levels to classify lots into 3 categories was investigated. The advantages of using dual sample acceptance levels over a single sample acceptance level was demonstrated for 2.27 and 9.08 kg samples.