chi-square
UK: ˈkaɪ skwɛə | US: ˈkaɪ skwɛr
n. a statistical method used to test the independence or goodness-of-fit of categorical data.
n. the distribution underlying this test, derived from the sum of squared standard normal deviates.
The term combines "chi," the 22nd letter of the Greek alphabet (χ), representing the distribution's mathematical foundation, and "square," referencing the squared deviations central to the test's calculation. Introduced by Karl Pearson in 1900, the name reflects the formula's reliance on χ² values, where observed and expected frequencies are compared via squared differences.
The researcher used a chi-square test to analyze voter preferences across age groups.
A significant chi-square result indicates a deviation from the expected distribution.
The chi-square statistic is calculated by summing (O-E)²/E for all categories.
Degrees of freedom critically influence the chi-square distribution's shape.
Chi-square assumptions include independent observations and adequate sample size.