Researchers developed a new test for comparing data with random censorship, similar to a well-known test by Karl Pearson. They found that their test is more accurate than a previous one in most cases, especially when the last data point is not too rare. The new test is better for some types of data, while the old test is better for others.