01.03.2013 Views

Applied Statistics Using SPSS, STATISTICA, MATLAB and R

Applied Statistics Using SPSS, STATISTICA, MATLAB and R

Applied Statistics Using SPSS, STATISTICA, MATLAB and R

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

310 7 Data Regression<br />

with s<br />

2<br />

= ∑ ∑<br />

2<br />

( d i1<br />

− d1<br />

) + ( d<br />

n − 2<br />

i2<br />

− d<br />

In the present case the computed t value is t * = –1.83 <strong>and</strong> the 0.975 percentile of<br />

t412 is 1.97. Since |t * | < t412,0.975, we accept that the residual variance is constant.<br />

Test of Fit<br />

We now proceed to evaluate the goodness of fit of the model, using the method<br />

described in 7.1.4, based on the computation of the pure error sum of squares.<br />

<strong>Using</strong> <strong>SPSS</strong>, <strong>STATISTICA</strong>, <strong>MATLAB</strong> or R, we determine:<br />

n = 414; c = 381; n – c = 33; c – 2 = 379 .<br />

SSPE = 1846345.8; MSPE=SSPE/( n – c) = 55949.9 .<br />

SSE = 34921109 .<br />

Based on these values, we now compute:<br />

SSLF = SSE − SSPE = 33074763.2; MSLF = SSLF/(c – 2) = 87268.5 .<br />

Thus, the computed F * is: F * = MSLF/MSPE = 1.56. On the other h<strong>and</strong>, the 95%<br />

percentile of F379, 33 is 1.6. Since F * < F379, 33, we do not reject the goodness of fit<br />

hypothesis.<br />

Detecting Outliers<br />

The detection of outliers was already performed in 7.3.2.1. Eighteen cases are<br />

identified as being outliers. The evaluation of the model without including these<br />

outlier cases is usually performed at a later phase. We leave as an exercise the<br />

preceding evaluation steps after removing the outliers.<br />

Assessing Multicollinearity<br />

Multicollinearity can be assessed either using the extra sums of squares as<br />

described in 7.2.5.2 or using the VIF factors described in 7.3.2.2. This last method<br />

is particularly fast <strong>and</strong> easy to apply.<br />

<strong>Using</strong> <strong>SPSS</strong>, <strong>STATISTICA</strong>, <strong>MATLAB</strong> or R, one can easily obtain the<br />

coefficients of determination for each predictor variable regressed on the other<br />

ones. Table 7.8 shows the values obtained for our case study.<br />

Table 7.8. VIF factors obtained for the foetal weight data.<br />

r 2<br />

2<br />

)<br />

2<br />

.<br />

BPD(CP,AP) CP(BPD,AP) AP(BPD,CP)<br />

0.6818 0.7275 0.4998<br />

VIF 3.14 3.67 2

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!