Personality is supposed to be independent of intelligence - Personality is a separate explanatory variable which can be seen after Intelligence is controlled-for.
Intelligence is primary as an explanation of behaviour - primary both historically, and because intelligence (very obviously) affects personality - but personality does not affect intelligence.
In other words, as a matter of routine - when measuring personality, one should also test for intelligence - and before looking at the effect of personality on behaviour, one ought to remove the effect of intelligence (by stratified analysis, preferably - i.e. creating narrow strata of IQ and only looking a personality effects within these strata - or else by some kind of regression).
However, much personality research is done on an already-intelligence-stratified sample - such as Psychology Students at Mudsville State University - in these situations the researcher can usually get-away-with missing out IQ testing and just evaluating Personality.
However, this does not apply to the pseudo-trait of Openness - which is often so sensitive to IQ differences that it varies even within strata such as the same class of the same college.
If Intelligence is controlled-for, then the effect of Openness disappears - because Openness is merely 'the personality type of intelligent people in Western-type societies' (but rather badly conceptualized).
While the other personality traits (C, E, A and N), which derive essentially from HJ Eysenck, are robust to IQ differences (especially in college populations which provide most of the subjects): Openness is not.
Openness merely a (weak) correlate of IQ (in Western Societies)... plus noise and cross-contamination from other personality traits (e.g. a little bit of Psychoticism/ Schizotypy).
Take home message: all research on so-called Openness is either ignorant, incompetent - or (usually) both.
This began as a comment on the Isegoria blog