Suitability of Input Data

Input data should be described in sufficient detail. That alone, of course, doesn t guarantee their quality but it allows for a judgment whether the data are suitable for the intended purpose. Input data could be weak for a variety of reasons  

Accuracy is the distance of a measured or estimated value to its true value. Inaccuracies (systematic deviations or bias) may arise from inadequate survey design and processing of the samples. Expert judgment is needed to assess the accuracy however, that implies that the origin and the extraction of the data is described (sampling design and conditions). 

Data must be representative of the scenario to be assessed. If the population and conditions under assessment are different from the data source, or the scope is broader than the data source, then caution must be exercised. In those cases it may be possible to demonstrate representativeness by bridging data or other supporting evidence. With regard to effects data, extrapolation from lab to field is often involved, thus introducing an additional degree of uncertainty. 

Where regression statistics are involved the relationship should not be extrapolated beyond the observed data range for the independent variable. 

This is particularly important if the linear regression model is used rather than a mechanistic model. 

---