Repeatability and reliability of an approach test to determine calves' responsiveness to humans: "a brief report"

A test aiming to measure calves' responses to humans ("approach test") was conducted in two steps: during the first step (approach phase) the calves' reaction to the appearance of an unknown person (withdrawal or not) was observed during milk provision, during the second step (touch phase) the calves' reaction to the outstretching of the arm and touching on the head was measured qualitatively on a scale from I (no withdrawal) to 4 (strong withdrawal). The repeatability of the withdrawal test was determined by performing the test twice with a 2-day interval on 150 calves. The reliability of the test was determined by comparing the responses of 22 calves to the approach test with their behaviour to humans during a longer test in an arena. For the approach phase, a high consistency was found between repetitions: calves that withdrew from the unknown person during the first observation, withdrew again the next time. For the touch phase, calves with either score I or 4 had the same score the following observation. However, responses of score 2 or 3 were not very consistent over time. The scores obtained in both the approach and touch phase of the test were strongly related with the calves' responses to a person in an arena: calves who withdrew from the unknown person at the approach test showed a longer latency to interact, interacted less often and shorter with the unknown person when observed in the arena than did calves who did not withdraw. In conclusion, the approach test can be considered to be repeatable and reliable. Therefore, it can be used to determine the animal's responsiveness to humans. (C) 2003 Elsevier BX All rights reserved.