Iterative Proportional Fitting
- Adjusts survey results to compensate for groups that are under- or over-represented in the sample.
- Compares sample demographics with known population characteristics and makes iterative adjustments.
- Helps survey results better reflect the entire population rather than only the surveyed individuals.
Definition
Section titled “Definition”Iterative proportional fitting, also known as raking, is a statistical method used to adjust survey data to match known population characteristics.
Explanation
Section titled “Explanation”The goal of iterative proportional fitting is to ensure that survey results accurately represent the entire population rather than only the sample of individuals who were surveyed. The method involves comparing the demographics of the sample to those of the known population and making adjustments to the survey results so they more closely reflect the overall population.
Examples
Section titled “Examples”Adjusting for under-representation
Section titled “Adjusting for under-representation”If a survey sample has a disproportionately low number of individuals from a certain racial or ethnic group, iterative proportional fitting can be used to adjust the survey results to account for that under-representation by comparing sample demographics to known population demographics and applying adjustments.
Adjusting for over-representation
Section titled “Adjusting for over-representation”If a survey sample has a disproportionately high number of individuals who are highly educated, iterative proportional fitting can be used to adjust the survey results to account for that over-representation by comparing the education levels of the sample to those of the known population and applying adjustments.
Use cases
Section titled “Use cases”Iterative proportional fitting is used to make survey results more accurate and reliable for informing decision making and policy development.
Related terms
Section titled “Related terms”- Raking