A test for the equality of the estimated coefficients in two linear regressions on two different data samples. The null hypothesis is that the coefficients in the two regressions are the same, and the alternative is that the coefficient on at least one explanatory variable is different. The test statistic involves the sums of squared residuals from three regressions: two separate regressions on two samples and one regression on a pooled sample. Under the null hypothesis the test statistic has an F-distribution. In time series the test is used to assess the stability of coefficients, that is, to assess whether there is a structural break in the model; the potential break point is assumed to be exogenous and must be specified ex ante. The test is not valid when the timing of the break depends on the data, or is endogenous.
|Reference: Oxford Press Dictonary of Economics, 5th edt.|