Towards a cost effective method for estimating disease infection rates in plants

Abstract

Pool-testing involves investigating more than one unit with a single test for cost effectiveness. Rather than testing an individual plant for the presence of a disease, plants are put into groups of more than one unit and subjected to a single test to determine the presence of a disease. Therefore, if the disease is absent a single test will be sufficient; otherwise, each individual plant is tested. The study used this procedure to estimate the infection rate of a disease. The method of maximum likelihood estimator for equal and unequal group sizes is discussed. For comparison with one-at-time testing procedure, properties of the constructed estimators have also been discussed. Empirical evidence via simulation of the optimal group sizes and asymptotic relative efficiency have been achieved by use of a statistical package. From the simulated results, it emerged that the proposed procedure has smaller asymptotic variance than one-at-a-time testing procedure. Therefore, the proposed procedure is more efficient than the usual one-at-a-time testing.