The indices bias and accuracy provide an objective indication of model performance. It has been shown that these factors were also valuable tools for evaluationthe performance of predictive models (Ross, 1996; Neumeyer et al., 1997; Dalgaard and Jrgensen, 1998; te Giffel and Zwietering 1999). A bias factor less than 1.0 indicates that the model is, in general, fail-safe. ResultsTable 2 show values of 0.99 for square root model, this predicted value being higher than the observed value. Eq. (2) (Arrhenius model) showed a bias factor of 1.11, indicating that the prediction resulted in a lower level of confidence.
Since the bias factor provides no indication of the average accuracy of estimates because under- and