Hypothesis Testing and Correlation
From Ian Stark
This lecture looks at multidimensional datasets and the idea of correlation: where values from two sets of data vary in some linked way. We can look for correlation in scatter plots, measure it with a correlation coefficient, and use hypothesis testing to detect whether there is evidence of some real connection. Beyond this analysis there are questions of causal connection — can we use statistical analysis to uncover when one thing causes another?
Link: Web page for this lecture