In psychometrics, the divergent validity of a survey instrument, like an IQ-test, indicates that the results obtained by this instrument do not correlate too strongly with measurements of a similar but distinct trait.
For example, if a test is supposed to measure suitability of applicants to a particular job, then it should not exhibit too strong correlation with, say, IQ-scores. Otherwise, the instrument is just another IQ-test.
The complementary concept is the convergent validity. Both are forms of the construct validity.