Under certain conditions, the two groups of data will meet certain properties when they are discussed separately, but once they are considered together, they may come to opposite conclusions. Simpson expounded this phenomenon in his published paper.
When people try to explore whether two variables (such as freshman admission rate and gender) are related, they will study them separately in groups. However, the dominant party in group comparison is sometimes the one who loses power in the overall evaluation. 195 1 year, this phenomenon was described by E.H. Simpson in his published paper before it was officially described and explained. Later, this paradox was named after him, namely Simpson paradox.