Validity and credibility

Discussion in 'CM1' started by AKA, Mar 14, 2019.

  1. AKA

    AKA Made first post

    In the section on model suitability (from the old CT4) the notes (and the IFoA answer to the specimen paper Q10) say that one must consider:

    A) Data/model/assumptions validity
    B) Data/results credibility

    What's the difference between them? I assume credibility is "is the data true or not?" - but what's validity then? If it's true how can it be invalid? Is it a relevancy level?
  2. pd97

    pd97 Made first post

    Validity of assumptions/data means how accurate is it to use a certain assumption. for example - if you use an assumption of the data distribution to be normal, you must run a hypothesis test (or some other means) to check if the data actually comes from a normal distribution.

Share This Page