Ameasurement of personality that produces observable, correct results, but that are not consistent would be considered:

I believe the answer is: valid, but not reliable

A result would be considered as valid if it is created by using proper logical thought or common way of reasoning. But, such result on personality would not be considered to be reliable because there are too many factors that could create a deviation within the data, and there is no accurate way to truly measure personality.

answer; ///relational devaluation;