Weighted Least Squares for Adjusting Efficacy Data with Inconsistent Spread (78 Patients)
- 972 Downloads
Linear regression assumes that the spread of the outcome-values is homoscedastic: it is the same for each predictor value. This assumption is, however, not warranted in many real life situations. This chapter is to assess the advantages of weighted least squares (WLS) instead of ordinary least squares (OLS) linear regression analysis.