Combining data from high and lower quality surveys

What is it about?

High quality surveys are expensive. To reduce costs, polling companies rely more and more on cheap data collected from the Internet surveys that are not statistically sound. We propose an approach that combines smaller and thus cheaper samples from the high-quality, statistically reliable surveys with larger samples from cheap Internet surveys. As a result, the overall cost of the data collection is smaller and the resulting analysis is still able to achieve desired accuracy as a larger, more expensive survey would do.

Featured Image

Photo by Elliott Stallion on Unsplash

Photo by Elliott Stallion on Unsplash

Why is it important?

The costs of collecting high-quality data through surveys and opinion polls are increasing and the typically used alternative methods often lack quality that may lead to incorrect conclusions.