Five methods of experimental reasoning distinguished by John Stuart Mill in his System of Logic (1843).
Suppose one is interested in determining what factors play a role in causing a specific effect, E, under a specific set of circumstances. The method of agreement tells us to look for factors present on all occasions when E occurs. The method of difference tells us to look for some factor present on some occasion when E occurs and absent on an otherwise similar occasion when it does not. The joint method of agreement and difference combines the two previous methods. The method of residues applies when part of E is explicable by reference to known factors, and tells us to attribute the "residue" to the remaining circumstances under which E occurs. The method of concomitant variation is used when E can be present in various degrees; if we identify a factor F, such as temperature, whose variations are positively or negatively correlated with variations in E, for instance, size, then we can infer that F is causally connected with E.