ABSTRACT
ABSTRACT Properly representing a population using survey data leaves analysts with a list of challenges. This chapter describes in detail how to work through one of the most common challenges: how to properly represent the population after the survey has been elded and the data have been collected. By using a poststratication technique called iterative proportional tting, or raking, in conjunction with known population demographics, the sampling weights can be adjusted for each individual respondent. Doing so essentially makes some respondents (those undersampled) represent more than one person, while others (those oversampled) will represent less than a person, forcing the data to represent the known demographics.
