Abstract. Title of the paper: Comparison of the conformation scoring and the body measurement based on four linear traits of Holstein Friesian cows
The goal was evaluate the similarity of the body scoring System and the body measurement kept generally for the best objective method. The target-traits (withers height, body depth, rump width at the gluteal tuberosity and body capacity) of Holstein Friesian heifers were taken down. Three-three groups (−extreme, medium, +extreme) based on the Standard deviation of the body measurements were created and compared in each traits, as well as phenotypic relationships (regression and correlation) were calculated. Taking the relationships into consideration it appeared that the body measurements do not stand in such a close connection with each other than it is suggested by the scoring.

The pair-wise correlation coefficients between the values of the same traits taken down by scoring and measuring Systems can by seen as low figures (r = 0.71, 0.50, 0.43 and 0.59, respectively) since we have the same traits. The partial correlation coefficients are even more less than the pair-wise correlation coefficients, and statistically confirmed only by the height at wither. Except the height at wither, there were no demonstrable connections in the regression of the same traits, just then, exclusively the true height at wither influenced the classifier during the scoring of all the traits.

The differences between the type groups were smaller in each traits estimated by scoring. From the analysis of the type groups it is obvious that the similarity of the two methods is the biggest in the height at wither. In the case of body depth, rump width and body capacity the classifier can distinguish only the "very little" and "very large" animals from the judged population while the intermediate types remained unrecognised. The average value of the individual identity between the two typing method was around 35%.