Reliability and validity are concepts that many students (and researchers!) have a hard time keeping straight.
Validity: the extent to which an empirical measure adequately reflects the real meaning of the concept under consideration. (how closely do you come to hitting the bull's-eye (presumably the concept you actually want to measure?)
Reliability: whether a particular technique, when applied repeatedly to the same object, will yield the same results each time.
With that in mind, some pictures:
|The "hits" are far from the bull's-eye (low validity);|
and all over the place (low reliability).
|Hits are closer to the bull's-eye (higher validity); but|
the hits are also all over (low reliability).
|Hits are FAR from the bull's-eye (low validity); but they are|
clustered in the same area-- the same answers are
received with each test (high reliability).
(more common with survey research (of course we strive
to hit the bull's-eye!))
|Hits are clustered in the bull's-eye-- high validity and high reliability.|