RE: Card sorting!!!

Well, since this thread continued in this vein, I guess my reason for intro=
ducing the "affinity diagramming" thread was because I saw a useful underlyi=
ng commonality that tended to add credence to the card sorting method. So fa=
r, following David's and Donna's process descriptions, I'm not seeing a vast=
 difference in either the method or the intentions of either -- although whe=
n I think about it, the intentions and assumptions will most certainly drive=
 the method. In my experience managing a small group of multimedia designer/=
developers, we met with our customer and used brainstorming and A-D'ing to g=
et a sense of the scope of the ultimate product. A-D'ing is considered "cons=
ensus building" only in the sense that it provides a common view and the abi=
lity for a team (including observers) to make a contribution. Our product en=
ded up being a set of video scenarios corresponding to types of non-complian=
ce with drug regimens used for asthma. From this basic understanding and cat=
egorizing of needs, we were able to do many things, including developing a t=
heme ("find yourself!") and even a cost model for the development effort. Of=
 interest perhaps was that we were able to understand the full scope of the =
product and come within budget pleasing both ourselves and the customer. We =
also created an excellent product that never really took off :) (That is a l=
onger story requiring a good deal of explaining).

Card Sorting strikes me as a near relative. I can't really see a conflict a=
nd my intention was only to raise an observation.

-----Original Message-----
From: Donna Spencer [mailto:[email protected]]=20
Sent: Wednesday, March 04, 2009 3:46 PM
To: User-Centred Design (Discussion)
Cc: User-Centred Design (Discussion)
Subject: Re: Card sorting!!!

There's no conflict, they just are different things, like dogs and cats are=
. They have the same core aspect of making groups from smaller pieces.

Card sorting is usually done with the users of a product. They may undertak=
e the sort as individuals or within small groups. Always more than one perso=
n/group will do the sort. You (the person who conducts the
research) analyse the results from the activity to look for similarities an=
d differences in approach that will help you make decisions about a project =
- usually to determine how content will be organised on something like a web=
site.

Affinity diagramming is usually done with a team of folks who work together=
. Its most common use is to analyse research data such as user research or u=
sability testing. An affinity diagramming session is done once, by a team, t=
o identify broad patterns and findings from lots of small data like research=
 observations. It doesn't have to be about building concensus, but about bui=
lding an understanding.

So they both have at their core that someone sorts granular data into group=
s. But they are done by different people for different reasons.

And the only reason the distinction matters is one of communication. If you=
 say to your boss - 'hey, I want to do a card sort', that means a different =
set of steps and outcomes compared to saying 'let's do an affinity diagrammi=
ng session'.

Donna

Tom Kerwin wrote:
> Hi,
>
> I must admit, I can't tell what the conflict is between the two technique=
s.
>
> Having looked into the links that William posted about Karen=20
> Holtzblatt's version of Affinity Diagramming, it seems to me that=20
> there's a key difference in the different groups of people the two=20
> techniques are most usefully applied to.
>
> It seems that Affinity Diagramming is used (at least by Karen's team)=20
> to gain understanding, consensus and buy-in from different parties=20
> within an organisation. Card Sorting (at least the way I've used it so=20
> far) is focused on gaining understanding of how users or customers=20
> outside the organisation understand conceptual groupings. Then you can=20
> compare the way the customers think with the way the staff (or expert use=
rs) think.
>
> I'm sure each technique could be used with the other group of people.=20
> I've done card sorting with staff before. I tend to include card=20
> sorting as part of user testing sessions, so it's generally done with=20
> an individual. It seems that affinity diagramming would need more of a=20
> focus group type setting.
>
> Having seen this info, I'm really keen to try affinity diagramming=20
> with difficult clients. I hope it can help avoid some of the political=20
> problems that often crop up in larger organisations.
>
> Cheers,
>
> Tom
>
>
>
>
> -----Original Message-----
> From: Peter Capell [mailto:[email protected]]
> Sent: 04 March 2009 15:05
> To: User-Centred Design (Discussion)
> Subject: RE: Card sorting!!!
>
> Hi,
>
> Ok, but in my original post on this comparison, I did mention=20
> something called "nominal group technique," which has to do with=20
> making the "category collapsing*" activity into a "private thing" (to=20
> be adjudicated after the fact). Anyway, I suspect that my raising the=20
> "affinity diagramming" idea may simply be an irritating comparison to=20
> those who feel enamored of "Card Sorting," so I think I'll drop this thre=
ad. Thank you for your reply.
>
>  Peter
>
> -----Original Message-----
> From: Diarmad McNally [mailto:[email protected]]
> Sent: Wednesday, March 04, 2009 10:01 AM
> To: User-Centred Design (Discussion); Peter Capell
> Subject: RE: Card sorting!!!
>
>
> Hi
>
> Generally, card sorting is carried out on individuals while affinity=20
> diagramming is a group activity. This means that AD results are=20
> subject to group dynamics to reach consensus while this is not so for=20
> CS as the responses are individual and it is the analysis which=20
> provides the 'consensus' which is reflective of the group.
>
> Diarmad
>
> --- On Wed, 4/3/09, Peter Capell <[email protected]> wrote:
>
>  =20
>> From: Peter Capell <[email protected]>
>> Subject: RE: Card sorting!!!
>> To: "Diarmad McNally" <[email protected]>, "User-Centred Design=20
>> (Discussion)" <[email protected]>
>> Date: Wednesday, 4 March, 2009, 2:49 PM Good Morning,
>>
>> Sorry, I failed to comprehend the _difference_ you are alluding to.=20
>> "Consensus" and "individual response" achieving it =3D "consensus."
>>
>>  Peter
>>
>> -----Original Message-----
>> From: Diarmad McNally [mailto:[email protected]]
>> Sent: Wednesday, March 04, 2009 4:03 AM
>> To: User-Centred Design (Discussion)
>> Subject: RE: Card sorting!!!
>>
>>
>> Morning,
>>
>> Another interesting difference between card sorting and affinity=20
>> diagramming is that the latter attempts to achieve consensus in a=20
>> group and is therefore subject to group bias (e.g. strong characters=20
>> leading the group etc.) while the consensus agreed through card=20
>> sorting is aggregated from individual responses.
>>
>> Regards,
>> Diarmad McNally
>> Interaction Design Studio
>>
>> t: (+44) 07808 297289
>> e: [email protected]
>> w: www.ixdStudio.com
>> Twitter: www.twitter.com/ixdStudio
>>
>>
>>
>> --- On Tue, 3/3/09, William Hudson
>> <[email protected]> wrote:
>>
>>    =20
>>> From: William Hudson
>>>      =20
>> <[email protected]>
>>    =20
>>> Subject: RE: Card sorting!!!
>>> To: "User-Centred Design (Discussion)"
>>>      =20
>> <[email protected]>
>>    =20
>>> Date: Tuesday, 3 March, 2009, 4:10 PM Peter -
>>>
>>> While I agree that there are some similarities between
>>>      =20
>> card sorting
>>    =20
>>> and affinity diagramming, I don't think that one
>>>      =20
>> lead to the other.=20
>>    =20
>>> Affinity diagramming comes from the KJ Method which
>>>      =20
>> was only invented
>>    =20
>>> in the 1960s. Philosophers, psychologists and
>>>      =20
>> information scientists
>>    =20
>>> (in various guises) have been categorising concepts
>>>      =20
>> for many hundreds
>>    =20
>>> of years before that. In fact, even the methods of
>>>      =20
>> cluster analysis
>>    =20
>>> and tree diagrams commonly used to process card sorts
>>>      =20
>> predate the KJ
>>    =20
>>> Method (although they came after Aristotle<g>).
>>>
>>> Also, there are some very important differences
>>>      =20
>> between card sorting
>>    =20
>>> and affinity diagramming. As it happens, I cover these
>>>      =20
>> in my web
>>    =20
>>> design course at CHI 2009 in Boston - see
>>>
>>>      =20
>> http://chi2009.org/Attending/AdvanceProgram/courseC119.htm
>>    =20
>>> Basically, card sorting is trying to investigate
>>>      =20
>> categories and
>>    =20
>>> terminology while affinity diagramming is much more
>>>      =20
>> about concepts,
>>    =20
>>> descriptions and the voice of the customer (at least
>>>      =20
>> as practised by
>>    =20
>>> Karen Holtzblatt and her colleagues). Anyone
>>>      =20
>> interested in affinity
>>    =20
>>> diagramming should have a look at Jared Spools web
>>>      =20
>> page on the subject
>>    =20
>>> (http://www.uie.com/articles/kj_technique/) and either
>>>      =20
>> a Karen
>>    =20
>>> Holtzblatt's books on contextual design.
>>>
>>> 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
>>>      =20
>> Wales.
>>    =20
>>> Registered number: 1895345. Registered office: 10
>>>      =20
>> Oxford Road,
>>    =20
>>> Abingdon OX14 2DS.
>>>
>>> Advances in Card Sorting - One Day Workshop
>>>  - London, 21 April 2009
>>> http://www.syntagm.co.uk/design/csadvances.shtml
>>>
>>>      =20
>>>> -----Original Message-----
>>>> From: Peter Capell [mailto:[email protected]]
>>>> Sent: 03 March 2009 15:19
>>>> To: UCD
>>>> Subject: RE: Card sorting!!!
>>>>
>>>> Hi,
>>>>
>>>> Card sorting is apparently a slight modification
>>>>        =20
>> to
>>    =20
>>> "affinity
>>>      =20
>>>> diagramming" about which there is much
>>>>        =20
>> (older)
>>    =20
>>> literature in the area
>>>      =20
>>>> of "meeting management." I think it is
>>>>        =20
>> a
>>    =20
>>> very powerful technique (when
>>>      =20
>>>> the group/team/users are cooperative). Often it
>>>>        =20
>> is
>>    =20
>>> coupled with first a
>>>      =20
>>>> session of brainstorming, which when conducted by
>>>>        =20
>> the
>>    =20
>>> rules is a very
>>>      =20
>>>> helpful adjunct. There are even permutations of
>>>>        =20
>>> brainstorming/card
>>>      =20
>>>> sorting (affinity diagramming) called
>>>>        =20
>> "nominal
>>    =20
>>> group technique" which
>>>      =20
>>>> allows for differences in personality type.
>>>>
>>>>  Peter
>>>>
>>>>        =20
>>> ---
>>> You are currently subscribed to ucd as:
>>> [email protected].
>>> To unsubscribe send a blank email to
>>>
>>>      =20
>> leave-44499-378.8bf1211fd4b7b94528899de0a43b9fb3@listman.syntagm.co.u
>> k
>>
>>
>>      =20
>>
>> ---
>> You are currently subscribed to ucd as: [email protected].
>> To unsubscribe send a blank email to
>> leave-44522-310.06eb61b839a0cefee4967c67ccb099dc@listman.syntagm.co.u
>> k
>>    =20
>
>
>      =20
>
> ---
> You are currently subscribed to ucd as: [email protected].
> To unsubscribe send a blank email to
> leave-44543-861.f9a40a4780f5e1306c46f1c8daecee3b@listman.syntagm.co.uk
>
>
> ---
> You are currently subscribed to ucd as: [email protected].
> To unsubscribe send a blank email to=20
> leave-44548-389.c86a7ee3d8ef0b551ed58e354a836f2b@listman.syntagm.co.uk
>
>  =20

--
Donna Spencer - Maadmob
[email protected]
02 6255 4993 / 0409778693
http://maadmob.com.au/
http://www.rosenfeldmedia.com/books/cardsorting/


---
You are currently subscribed to ucd as: [email protected].
To unsubscribe send a blank email to leave-44578-310.06eb61b839a0cefee4967c=
[email protected]

---
You are currently subscribed to ucd as: [email protected].
To unsubscribe send a blank email to leave-44590-427.18997733ec258a9fcaf239=
[email protected]