Auteur: Becker T.H., Blume L., Kayser M. and Isselstein J.
Jaar van uitgifte: 2015
Questionnaires are a frequently used instrument to analyse the productivity of farms. As surveys might include wrong or incorrect data, there is a need for validity testing. A validity test aims at generating an adjusted, reliable data-set with fewer outliers. Large data sets require an automated approach. We conducted a survey on 47 German dairy farms to evaluate the role of grassland in milk production. The farms are located all over Germany with a focus on Lower Saxony and Hesse. In a first step, we developed a generally applicable validity test for assessing the milk yield directly related to grassland. Several simple, directly measurable parameters were defined which correlate with important parameters with a high error rate. These relations were put into a formula and applied to the data set. We found that out of the 47 data sets three had to be excluded from further analysis because of large deviations from the defined confidence limits. The experience with the validity test did not only result in a more reliable data set, but helped to optimize the questionnaire for future surveys. The farms in this survey produced 4,916 l grassland-milk ha‑1.