SUMMARY: Asking about typical or specific instance

Thanks to all who took the time to respond to my original question (in
the subject line above and repeated in full at the foot of this note).

There seems to be general agreement (but not consensus!) that it is best
to get users to think about a specific situation, rather than a
'typical' instance, which may be too heavily filtered or vague. A
concern was expressed that an individual instance may not be 'typical',
but since I am hoping to have a large number of respondents, taking an
average across recent, specific instances should give me the kind of
overall picture I am looking for. (I hasten to add that the situation
would be quite different if I was doing qualitative or ethnographic
research.)

The individual messages I received are below.

Regards,

William Hudson
Syntagm Ltd
Design for Usability
UK 01235-522859
World +44-1235-522859
US Toll Free 1-866-SYNTAGM
mailto:[email protected]
http://www.syntagm.co.uk
skype:williamhudsonskype 

Syntagm is a limited company registered in England and Wales (1985).
Registered number: 1895345. Registered office: 10 Oxford Road, Abingdon
OX14 2DS.

Advances in Card Sorting - One Day Workshop
 - London, 5 October 2009
http://www.syntagm.co.uk/design/csadvances.shtml 

-----

The specific event, initially, and then use that to establish whether
the
event was typical or not.

Using the specific event first gives you better and more accurate data,
and
then places the general information into a context. 

I often use a trilogy: e.g. when asking about work, I'd ask them to
think
about (or even better, to bring with them to talk about) an easy piece
of
work, a typical piece of work, and then something that is particularly
important to them such as something they really don't want to get wrong
for
some reason. This is highly illuminating and is enormously more useful
than
just asking them about their work in general. 

I found this book really helpful in understanding what's going on:
Thinking About Answers: Application of Cognitive Processes to Survey
Methodology 
Seymour Sudman, Norman M. Bradburn, Norbert Schwarz

http://www.amazon.co.uk/Thinking-About-Answers-Application-Methodology/d
p/0787901202/ref=sr_1_1?ie=UTF8&s=books&qid=1243592105&sr=1-1 

(Actually, pretty much anything by that trio is worth reading. They
change
the order of their names on what they write).

Cheers
Caroline [[email protected]]

-----

William --
If you have limited time, ask about the most recent event and when it
occurred -- this gives you some information on frequency of the
experience as well as the user's attitude. Otherwise ask for both.
Bob Warren
[email protected]

-----

What a great question!

So, my personal priorities are:

- Specific event within the last day
- Specific event within last 7 days
- If more than 7 days since the last event - probe whichever specific
event they recall best
- Lastly, events in general

Dave [[email protected]]

-----

Sometimes their memory is best triggered by a critical incident. "can  
you tell me about a time something went wrong ... ? " or some  
variation. Otherwise , have them frame it as a story.
Michele [[email protected]]
Sent from my iPhone

-----

Hello,
If the goal is to gauge user's attitude to the events in general, then
I'd ask them to think about them in general. Asking about a recent event
may skew the results since it may been an outlier.

Aras Kannu | Principal Architect, User Experience | Infor | office:
678-319-8983 | [email protected]

-----

Speaking as someone trained as a therapist, always go for the specific.
When you ask for summaries, you often get considered, filtered, logical
answers. That might sound good, but they are often far from the truth of
the situation - idealized, details left out, roadblocks and emotions
minimized, etc. It's the same reason why conducting a contextual inquiry
is better than a structured phone interview or asking what prompts you
to do X is better than asking why you do X. 

- Jenny [[email protected]]

-----

Thanks William,

I've found that in almost all cases it's an easier question to answer  
the more specific the focus.

So, 'If you think about the last time you contacted support for help,  
how well did that encounter work for you?"
is easier to answer than 'when you make support calls, in general how  
is the experience'?

Of course, it might also be good to follow up with "was your last  
experience similar to other experiences that you had?"

Ron [[email protected]]

-----

Hi, William:
Regarding your post to CHI-Web about specific or typical instances. In
my opinion, each has its value. If you are looking for numeric scores
(e.g. rate this service on a scale of 1 - 7) then looking for typical
instances is the way to go. If you are trying to capture qualitative
data, such as "How did you feel about this service", asking the person
to reflect on specific instances will often evoke richer information,
more descriptive than asking about a general impression.

I hope this helps!

-matt

[email protected]
Vision & Logic
   781-259-9498

-----

Hi William,

I understand attitudes to be a manifestation of beliefs which are
expressed verbally and behaviourally in response to a specific context
or trigger. I recommend placing users in a specific context to gauge
'real' reactions or use the critical incident technique by asking
him/her to recall an event, occurring recently or otherwise, but one
which has stuck with them for some reason.

This is not to say that a more general question to illicit attitudes
wouldn't trigger a response based on a user's memorable event. I believe
the risk of being general is getting more general responses.

Best regards
Barry [[email protected]]


> -----Original Message-----
> From: ACM SIGCHI WWW Human Factors (Open Discussion) [mailto:CHI-
> [email protected]] On Behalf Of William Hudson
> Sent: 29 May 2009 10:31
> To: [email protected]
> Subject: Asking about typical or specific instance
> 
> All -
> 
> If you are trying to gauge users' attitudes to events, is it better to
> ask them to think about a specific event (the most recent one, for
> example) or their attitudes about this kind of event in general (or
> both)?
> 
> Reply to me and I will summarize for the list.

    --------------------------------------------------------------
        Tip of the Day: Forward out-of-office replies to
                    mailto:[email protected]
     CHI-WEB: www.sigchi.org/web POSTINGS: mailto:[email protected]
              MODERATORS: mailto:[email protected]
       SUBSCRIPTION CHANGES & FAQ:  www.sigchi.org/web/faq.html
    --------------------------------------------------------------