Creating Survey Weights in R Using Census Data
Published:
In survey research, the composition of a sample may differ notably from the population being modeled across important characteristics (e.g. race, age, education, party identification). These sampling errors often reflect systematic bias which can pose a threat to accuracy and the researcher’s ability to make inferences using the data, especially if the error is correlated with the variables of interest.