Method and apparatus to generate audience measurement data from population sample data having incomplete demographic classifications

Abstract

Methods and apparatus to generate audience measurement data from population sample data having incomplete demographic classifications are disclosed. An example method includes improving an accuracy of the first impressions counts by: determining a database proprietor frequency value for ones of demographic groups for a first one of time segments; calculating a distributor-level unique audience adjustment factor for ones of the demographic groups for the first one of the time segments; determining adjusted unique audience sizes based on the distributor-level unique audience adjustment factor for ones of the demographic groups for the first one of the time segments; for ones of the demographic groups for the first one of the time segments, calculating adjusted impressions counts by determining a product of the adjusted unique audience sizes and the database proprietor frequency value; determining a scaling factor for the first one of the time segments by determining a ratio between site census impressions counts and a summation of the adjusted impressions counts associated with the ones of the demographic groups for the first one of the time segments; applying the scaling factor to the adjusted impressions counts associated with the ones of the demographic groups for the first one of the time segments to generate adjusted and scaled impressions counts; and determining an adjusted and scaled duration for ones of the demographic groups for the first one of the time segments by multiplying the adjusted and scaled impressions counts associated with the first ones of the demographic groups by a corresponding time value associated with the first one of the time segments.