4.1.2 Durbin-Watson test
The Durbin-Watson test is conducted in statistics to check for
autocorrelation existing between the residuals in a linear regression
analysis(Chen, 2016). A positive autocorrelation value between 0 to 2
means that the model fits properly(Hepple, 1998). A value less than 0
and more than 2.5 is a cause for concern as it would mean too much of
outliers in the data or model overfitting. In this study, the Durbin
Watson value was found to be 1.65 which shows that model fits well and
accurately(Chen, 2016; Happé & Frith, 2006; Hepple, 1998).