This model was more appropriate than a random-effects ordinary least squares (OLS) regression, as confirmed by highly significant F tests and Hausman tests ( p s < .001).
1
—
—
This model was more appropriate than a random-effects ordinary least squares (OLS) regression, as confirmed by highly significant F tests and Hausman tests ( p s < .001).