next up previous
Next: Wilcoxon Estimation Up: Regression Previous: Regression

Least Squares (LS) Estimation

Least squares regression is the default analysis. The button for LS regression should already be checked. Most of the analyses in this course use the intercept but you may drop the intercept term (force the line to go through the origin) by unchecking the "Include an intercept" button. For the Height-Weight data using "Weight" as response and "Height" as predictor we obtain the following output.
Residuals: 
    Min      1Q  Median      3Q     Max  
-22.400  -8.938  -1.608   7.044  27.406  
 
Coefficients: 
            Estimate Std. Error t value Pr(>|t|)    
(Intercept) -272.808    136.113  -2.004  0.07998 .  
Height         6.403      1.814   3.530  0.00773 ** 
--- 
Signif. codes:  0  `***'  0.001  `**'  0.01  `*'  0.05  `.'  0.1  ` '  1  
 
Residual standard error: 15.39 on 8 degrees of freedom 
Multiple R-Squared: 0.609,	Adjusted R-squared: 0.5602  
F-statistic: 12.46 on 1 and 8 degrees of freedom,	p-value: 0.007729
Hence the estimated LS regression line is Weight = -272.8 + 6.4*Height and it accounts for 60.9% of the variation in weight.


2000-08-21