From: Rolf Michelsen <Rolf.Michelsen@delab.sintef.no>
Date: Thu, 25 Aug 94 01:14:39 PDT
To: Lucky Green <shamrock@netcom.com>
Subject: Re: Anonymous questionnaires
In-Reply-To: <199408241524.IAA26120@netcom7.netcom.com>
Message-ID: <Pine.3.89.9408250957.B1424-0100000@svme.er.sintef.no>
MIME-Version: 1.0
Content-Type: text/plain


On Wed, 24 Aug 1994, Lucky Green wrote:

> My partner is a participant in a long term psychological study. I have been
> asked to fill out a questionnaire to aid in this study. Some of the
> questions address issues that I would never answer non-anonymously. After
> speaking with the research director, I ended up with the following problem:
> is there a way that would allow the institute to

Interesting problem.  The suggestion (by whom?) to use two envelopes is 
useable as long as there is a trusted party involved.  This is very 
similar to the way such surveys are performed in Norway:  Each questinare 
has a random number on top and a trusted party is able to link numbers to 
participants' names and check who has answered and so on.  Researchers 
only see numbers and not names.  The problem is that the trusted party is 
often very close to the researchers using the survey...

Here is a suggestion for an electronic solution based on anonymous 
electronic coins:

You fill in your form and submit it electronically to the survey 
organizer.  The organizer acknowledges your form by giving you a blind 
signature much in the same way as a withdrawal in a Chaumian electronic 
cash system.  Later you unblind the signature and send it to the 
organizer together with name and adress to be registered as a 
participant.  The blind signature prevents linking of your name to the 
returned form but still proves that you have returned a form.

> 1. Correlate my answers to the answers of my partner.

Not directly provided by this simple solution, but the suggestion made by 
Stephen D. Williams to link you and your partner by writing down the same 
random number on the returned forms can be used.  There are other ways to 
link anonymous transfers too, but I won't come into that now...

> 2. Verify that I have indeed sent in a filled out questionnaire (and send
> me a check for participating).

OK.  They get your name together with the unblinded "coin" to prove your 
participation.

> 3. Allow a supervisory agency, such as the U.S. Department of Health and
> Human Services, to verify that the researchers did not just make up all the
> data - that is to allow an audit.

Same as above.

> 4. Protect my privacy by making it impossible to correlate my name to the
> answers given.

OK due to properties in the anonymous cash schemes.

The problem with this seemingly simple approach is that it requires an 
anonymous online connection between you and the survey organizer.  
Confidential and/or anonymous channels does not seem to be "in" among 
network providers today... :-(


-- Rolf


----------------------------------------------------------------------
Rolf Michelsen                           "Nostalgia isn't what it
Email: rolf.michelsen@delab.sintef.no     used to be..."
Phone: +47 73 59 87 33                   
----------------------------------------------------------------------