That depends, of course, on context, which you did not provide.
E.g. How do the PPV and NPV compare to other tests for the same issue?
As a screening test acceptable, but needs to know what you are screening for?
With the numbers for PPV and NPV you show, the disease prevalence must but larger than 12.5% and it seems likely that it is much larger than that.
That does not seem like a screening test to me.
May I suggest that the question whether to call a test statistic high or moderate is not terribly relevant, but the the question should be: is it good to use this test compared to the alternative?
The alternative is perhaps a gold-standard test that you may not want to use because it is costly. Then the question can be parsed into: What are the disadvantages of a false negative and of a false positive result, respectively? And are those disadvantages small enough to decide to save costs by not using the gold-standard test?
The alternative may be something else than a gold-standard test but it still seems best to focus on the advantages and disadvantages (incl. costs, if you will) rather then whether to call something arbitrarily high or moderate.
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning