Resumen

SANTOS FUENTES, Eric Eduardo; MONTESINOS LOPEZ, Osval Antonio  y  ANDRADE ARECHIGA, María. Sample sizes that ensuring accuracy to estimate prevalence of plants under inverse sampling. Rev. Mex. Cienc. Agríc [online]. 2016, vol.7, n.7, pp.1499-1512. ISSN 2007-0934.

The detection of a rare or scarce event (with low prevalence ≤ 0.1) in the design of agricultural experiments of a population consumes many resources. Therefore, one resorts to the inverse sampling (negative binomial) which consists of a series of tests with binary response (presence or absence) in which not stop sampling until a predetermined individuals with the trait of interest number. Therefore a method is proposed to calculate the required sample size (number of positive units) under inverse sampling ensures accurate estimated proportion because it ensures that the amplitude (W) of the confidence interval (IC) will be equal to, or more narrower than, the desired amplitude (ω), with a probability y (assurance level). Given the complex and laborious process of estimating both the sample size and parameters of interest (proportion, variance, standard deviation, total and confidence intervals for the proportion and total) free software is proposed for inverse sampling under the approach of accuracy in the estimation of parameters that automates the calculation of sample sizes and parameters of interest. In addition, the software provides a graphical, easy, safe and user friendly interface. The using the formula proposed for ensuring a probability y (assurance level > 0.5) fixed a priori accuracy is met IC is recommended. Which produces more accurate study performed interest.

Palabras llave : confidence interval; low prevalence; parameter estimation.

        · resumen en Español     · texto en Español | Inglés     · Español ( pdf ) | Inglés ( pdf )