This study aimed to verify if the selection is influenced by partial observation of laying instead of observation of total daily posture. On the first state of analyses, through a Bayesian model, it was considered as analyzed characteristics the parameters of laying probability curve in a model of random effects with function of logistic connection. On the second stage of the procedure, it was obtained the estimatives of (co)variance components and genetic parameters. It was observed an overestimation of the production mean curve for the total sample in systematic collections (collected once every 14 days, 1 or 2 times every 7 days) and partial (on 15, 30, 45, 60, 75 and 90 consecutive days). For the parameter that represents the beginning of laying (a), the heritability values were 0.54; 0.14; 0.12 and 0.07; and for the parameter that represents the laying evolution (β), they were; 0.36; 0.37; 0.40 and 0.47, respectively, in the systematic and total collections. In partial and total collections, respectively, the values were 0.04; 0.03; 0.06; 0.12; 0.17 and 0.25; 0.25; 0.33; 0.33; 0.32; 0.41 and 0.51. In the systematic collections, as heritability values for the parameters that represent the beginning of laying decreases, the other increases; however, as the sample size increases in partial collections, the values increase for both parameters. Within the top 100 animals classified according to a mediated selection index, the correlation among the scores of selected animals in the sub-samples and the total sample is low (from -0.13 to 0.13) indicating that the used selection index is affected.
animal selection; Bayesian analysis; correlated binary data; genetic parameters; nonlinear curves of production
