INTRODUCTION
MATERIAL AND METHODS
Data set
Penalty method
C1: average FS according to the number of inseminations. P = nx, where nx is the average number of FS days at the number of times till the last insemination plus one.
C2: constant number of days (359), derived from the highest expectation of FS in the records, P = nm, where nm = 359.
C3: maximum number of FS days to each insemination. P = nn, where nn is the maximum number of FS days to the nth insemination.
C4: average of nx and nm, where nx is the average number of FS days at the number of times till the last insemination.
C5: a constant number (21 d) was added to the highest number of FS days in each contemporary group. P = ng+ 21, where ng is the maximum number of FS in each contemporary group.
Statistical analysis
Bootstrap method
Generate a resampled data set 100 times from the original (non-censored) data.
Choose 8.7% of missing records within each resampling data set by the random censoring scheme.
Estimate genetic parameters and predict EBV for each data set.
Calculate the average of genetic parameters and predicted EBV.
Compare the penalty methods by bias and MSE of heritability and correlation of EBV with a non-censored data set.
Repeat steps 1 to 5 for three other percentages (5%, 10%, and 15%) to assess the influence of different proportions of missing records on the inference.