How to Create a Weighted Survey

Keep survey results organized to prevent clerical errors during the reporting process.
Keep survey results organized to prevent clerical errors during the reporting process. (Image: Jupiterimages/Comstock/Getty Images)

A survey sample may not accurately portray the greater population for which it is meant to provide information due to overrepresented or underrepresented demographics. To preserve the integrity of a survey, statisticians employ an approach known as "weighting" responses. Weighting responses is a mathematical approach to presenting survey results that scale responses according to statistics of the population. Yo will need to research your target population to determine the demographic distribution of the population, e.g., the number of women..

Divide the number of individuals in a category, such as men, by the number of individuals in your target population , such as a town. This will always provide a decimal less than 1. For example, 500 men in a town of 1,000 will equal 0.5.

Repeat Step 1 for your survey demographics to create decimals of each represented group's prevalence. For example, 40 men out of 100 responses would equal 0.4.

Divide the resulting population decimal by the resulting survey decimal. For the examples outlined above, this would mean a division of 0.5 by 0.4, with a result of 1.25. This number is the "weight" that will be applied to all results cast by men within your survey.

Multiply your newly obtained "weight" by the results of your original survey to obtain a representative statistic. Using the examples above, this would mean that, given 20 of 40 men responding "yes" within the survey, the representative results would display 25 men responding "yes" instead.

Related Searches


Promoted By Zergnet


You May Also Like

Related Searches

Check It Out

Are You Really Getting A Deal From Discount Stores?

Is DIY in your DNA? Become part of our maker community.
Submit Your Work!