The article introduces new ways to estimate the average of a group by using information about the proportion of that group with a specific trait. The researchers calculated the bias and error of these estimators under a simple random sampling method. They then tested their methods on real-world data to show how well they work in practice.