|
When you are first learning nonlinear regression, you can skip the choices on this Compare tab. But don't forget to come back and learn how Prism can help you compare models and datasets, as your scientific goals will often include comparing models.
|
When fitting biological data with regression, your main objective is often to discriminate between different models, to ask if an experimental intervention changed a parameter, or ask if the best-fit value of a parameter differs significantly from a theoretical value. Learn more about these three kinds of comparisons. Your choice, of course, has to be based on your experimental goals.
Prism can perform the comparison using two alternative methods: the extra sum-of-squares F test, and using Akaike's information criteria. Use these guidelines to choose:
| • | In most cases, the two models will be 'nested'. This means that one model is a simpler case of the other. For example, a one-phase exponential model is a simpler case of a two-phase exponential model. Either the F test or the AICc method may be used with nested models. The choice is usually a matter of personal preference and tradition. Basic scientists in pharmacology and physiology tend to use the F test. Scientists in fields like ecology and population biology tend to use AICc. |
| • | If the models are not nested, then the F test is not valid so you should choose AICc. Note that Prism does not enforce this. It will calculate the F test even if the models are not nested, but the results won't be useful. |
|
