The t statistic is a measure of how extreme a statistical estimate is. You compute this statistic by subtracting the hypothesized value from the statistical estimate and then dividing by the estimated standard error. In many, but not all situation, the hypothesized value would be zero.
|