Yate’s correction, also known as Yate’s chi-squared test, is used to test independence of events in a cross table i.e. a table showing frequency distribution of variables. It is used to test if a number of observations belonging to different categories validate a null hypothesis. It is a correction made to chi-square values in a binomial frequency distribution table.
It considers that the discrete probability of these frequencies is very near to continuous chi-squared distribution.However, to reduce the error in the assumption, a correction for continuity was suggested by Frank Yates. It is done by reducing the difference between each observed value and its expected value in a binomial frequency table by 0.5. These tests are commonly used when expected frequencies are less than ten.
Formula : XYates2 = ∑(│Oi –Ei│- 0.5)2/Ei
XYates2 = 0.0015
These tests are commonly used in diagnostics, environmental sciences, biology, etc.