Combining experiments with systematic errors

We consider fits to two or more datasets for which results from the same experiment share a common systematic uncertainty in addition their individual statistical errors. This is important extracting maximum information set of similar but different experiments (or under conditions) analysing datasets, as happens at LHC and other particle colliders. Two techniques can be used: one involves inverting full covariance matrix introduces extra parameters uses only diagonal terms. show, completely general fit, that an additive they are principle equivalent even though practice details differ. For multiplicative error fit parameter if factor applied data points not it function: former leads biassed estimates latter avoids them.