Bootstrap Methods & Permutation Tests

In this lesson, you'll learn what statisticians do when they need to analyze data that does not meet the strict requirements for the tests covered so far in this course. In our previous lessons, the tests have required Normal distributions and proper sample sizes. But sometimes these conditions simply cannot be met. For instance, some investigations involve collecting data on human subjects with serious illnesses where it would be simply impossible or unethical to keep sampling in order to meet all the criteria for traditional statistical analysis. Other times, the sample sizes are limited due to expense or the availability of subjects to study. Rather than simply discard the data or abandon the study, statisticians have gotten around the problem by using the power of computers to simulate the necessary conditions in a way enables them to make reliable estimates and draw valid conclusions. While these methods make for a far more complex analysis, they allow researchers to investigate important data sets they could not have otherwise analyzed.