Precise values for absolute receiver antenna phase centre corrections (PCC) are one prerequisite for high-quality GNSS applications. Currently, antenna calibrations are performed by different institutes using a robot in the field or in an anechoic chamber. The differences between the antenna patterns are significant and require a sound comparison concept and a detailed study to quantify the impact on geodetic parameters, such as station coordinates, zenith wet delays (ZWDs) or receiver clock estimates. Furthermore, a discussion on acceptable pattern uncertainties is needed. Therefore, a comparison strategy for receiver antenna calibration values is presented using a set of individually and absolutely calibrated Leica AR25 antennas from the European Permanent Network (EPN), both from the robot (Geo++ company) and from the chamber approach (University of Bonn). Newly developed scalar metrics and their benefits are highlighted and discussed in relation to further structural analysis. With our metrics, properties of 25 patterns pairs (robot/chamber) could be used to successfully assign seven individual groups. The impact of PCC on the estimated parameters depends on the PCC structure, its sampling by the satellite distribution and the applied processing parameters. A regional sub-network of the EPN is analysed using the double difference (DD) and the precise point positioning (PPP) methods. For DD, depending on the antenna category differences in the estimated parameters between 1 and 12 mm are identified also affecting the ZWDs. For PPP, the consistency of the observables, i.e. potential differences in the reference point of carrier phase and code observations, additionally affects the distribution among the different parameters and residuals.