Data

In confined systems where lactating cows were subdivided into multiple groups, the sample was proportionally defined according to the size of each herd. Data collection followed the Welfare Quality® protocol for dairy cattle and was structured around its four principles: Good Feeding, Good Housing, Good Health, and Appropriate Behaviour. Individual indicators were recorded at cow level, while herd-level measures were obtained from group observations, farm records, and interviews with farm managers.

Evaluations were carried out consistently between 07:00 and 15:00, always beginning after the first milking. Each farm was visited once, and cows were observed while standing freely in their usual environment, without the use of physical restraint or handling. To avoid duplicate assessments, the identification number of each cow was recorded prior to evaluation.

Assessors were blinded to the clinical history of the herds to minimize the risk of selection bias.