Even if the experiment has been carefully designed and the trials meticulously conducted, the modeler needs to appraise the accuracy of the data before attempting to fit the model. How were the data collected? What is the accuracy of the measuring devices used in the collection process? Do any points appear suspicious? Following such an appraisal and elimination (or replacement) of spurious data, It is useful to think of each data point as an interval of relative confidence rather than as a single point. This idea is shown in Figure 3.4. The length of each interval should be commensurate with the appraisal of the errors present in the data collection process.