Two basic approaches to statistical analysis included in many statistical packages are ordinary least squares (OLS) and maximum likelihood estimation. This entry considers the implications and practices involved in OLS. OLS employs a procedure most often associated with typical statistical procedures and corresponds to many common techniques in use (correlation, t-test, mean). The term least squares corresponds to the idea that the best value of estimation involves a parameter that minimizes the value of the sum of squared error or deviation. For example, one definition of the mean involves the value for a set of data where the sum of squared deviation from that value is the smallest. The term or estimate often becomes evaluated as the best fit or most accurate representation ...
Looks like you do not have access to this content.