A Study of Sample Size in Macadamia Nuts (Macadamia Integrifolia)

Abstract

Replicated samples of 10, 20, 30, AO, and 50 in-shell macadamia nuts were set up to determine optimum sample size needed to determine the number of nuts necessary to obtain acceptably reliable values for percent grade 1 kernels, percent kernel, and average shell diameter. Average shell diameter was found to be most variable with an average coefficient of variation of 150 percent for the 5 sample sizes. A 30 nut sample was necessary for +3 percent accuracy at the 95 percent probability level. The lowest average coefficient of variation found in the experiment was 2.4 percent for percent kernel. A 30 nut sample was required for accuracy of +2 percent at the 95 percent probability level. The average coefficient of variation for percent grade 1 kernels was 7.3, and a 100 nut sample was required for accuracy of +5 percent, at the 95 percent probability level.