The entire Chapter 6 titled “Rheological Models for Unified Curves” in the book A. V. Shenoy and D. R. Saini, Thermoplastic Melt Rheology and Processing, Marcel Dekker Inc., New York (1996) is dedicated to rheological models for polymer processing.  The criterion of the choice of the model depends on the shear rate range of interest, which in turn depends on the polymer processing operation that is being considered.  Each polymer processing operation subjects the polymer to a different range of shear rate, and the model that best fits would be the one that covers the particular range of interest.

In the above book, the model parameters for the Carreau Model are given in Table 6.5, for the Ellis Model given in Table 6.6, for the Ostwalde-de Waele Power-law Model given in Table 6.7 and for the General Rheological Model given in Table 6.8 for different generic-types of thermoplastics including Acrylonitrile Butadiene Styrene (ABS).

The method of your measurement is key in your model choice, because it determines your shear rate or frequency regime. In most kinds of piston rheometers the power-law does a good job. In plate-plate or cone-plate measurements you might find a leveling-off from log-linear behavior. Then, the first choice would be Carreau model, because its parameters are linked to the power-law parameters. Ergo you can fit the high shear data with the power-law and thus easily find a fit for the remaining value, which is the zero-shear viscosity. I personally would not rely on parameters you are able to find in books, because I measured different values even for the same resin in repeated measurements of different grades.