There's no alternative to a test chart and tripod (under different lighting conditions if necessary) for being certain of this kind of thing. I use the one at the bottom of
this page, but anything would do that indicated unequivocally where the focus actually is, compared with where the camera thinks it is.
Ideally it needs to be done with a different body, too, to isolate the problem to the lens or the camera.
I have a DA 17-70 that's just gone in for repair (out of warranty, sadly, since I acquired it used), since it back-focuses more than the K-5 adjustment allows. It's gone in with printed test shots, showing the results on two bodies (K-5 and K10D) and also the result with a lens that's perfect.