Concerning the question of publication and confidentiality:
It depends in large measure on the specifics of what is being requested
and/or promised.  In most survey studies the confidentiality of the
respondents would be expected in any case.  In many other cases the
results the client is interested in would not be of academic value
anyway.  It is usually possible to reserve the right to publish
various parts of the data, or to publish discussions of the methodology,
without upsetting the client.  And it is usually possible to negotiate
a procedure in which the client has the right to review manuscripts
before publication, at least with the right to comment and perhaps
with the right to prevent the publication of trade or commercial secrets.
It may also be possible to get the client to release the results after
some period of time that would protect their commercial competitiveness
but not unduly restrict academic discussion.  Talk it through with the
client and the investigator to see where the sensitive issues really are.

Stuart Ross
California State University Fullerton