Survey Raking: An Illustration

Analysing survey data can be tricky. There’s often a mismatch between the characteristics of the survey respondents and and those of the general population. If the discrepancies are not accounted for then the survey results can (and generally will!) be misleading. A common approach to this problem is to weight the individual survey responses so that the marginal proportions of the survey are close to those of the population. Raking (also known as proportional fitting, sample-balancing, or ratio estimation) is a technique for generating the required weights.