> In our recent studies, only two of the sites we tested had any
> frames (Fidelity & Disney). While neither of these sites did
> spectacularly, the frames didn't seem to be the issue.
One of the reasons you didn't find problems with frames may be the
artificial circumstances of a lab study. You wouldn't see any
problems with bookmarking or with emailing URLs to friends and
colleagues (one of the main ways in which good sites spread).
And then, of course, these sites may be among the more reasonable
ones in use of frames. It's one of these "ropes to hang youeself by" kinds