This paper looks at the variation in performance of a high density connector including insertion depth, return path via location and solder variations. Model validation using measurements is only possible if the reference plane locations are common between the measurement and model. Measurement details including the impact of calibration structure on correlation are presented.