Abstract: Techniques are provided for generating adjustment factors, on a per-demographic-group basis, to compensate for errors in audience verification statistics caused by co-usage of devices. The adjustment factors are based on a comparison of (a) per-group counts produced by applying one audience verification approach to a set of exposures, to (b) per-group counts produced by applying another audience verification approach to the same set of exposures. For example, the first per-group counts may be produced under the assumption that the user to whom content is exposed is the owner of the device, and the second per-group counts may be produced by obtaining demographic information directly or indirectly from a user at the time the user is exposed to the content.

Abstract: Techniques are provided for generating adjustment factors, on a per-demographic-group basis, to compensate for errors in audience verification statistics caused by co-usage of devices. The adjustment factors are based on a comparison of (a) per-group counts produced by applying one audience verification approach to a set of exposures, to (b) per-group counts produced by applying another audience verification approach to the same set of exposures. For example, the first per-group counts may be produced under the assumption that the user to whom content is exposed is the owner of the device, and the second per-group counts may be produced by obtaining demographic information directly or indirectly from a user at the time the user is exposed to the content.