[Sigia-l] card sorting: dealing with multiple placements

[Eric Scheid]

On 1/6/03 9:03 AM, "Boniface Lau" <boniface_lau at compuserve.com> wrote:

>> I've done "card sorts" with users many times, and never have we used some
>> mechanical device for processing the cards into some arrangement determined
>> by intrinsic qualities represented on those cards.
>> 
> First, the concept of sorting. Sorting is about order. But IAs use card sort
> for its grouping effect. Thus, I said IAs use an ordering tool for grouping
> purpose.
 
Do you mean 'order' as in 'chaos', or as in 'sequence'? By what you've
written so far I'm assuming the latter. I'd agree with the former, but not
necessarily the latter. Grouping, without internal sequencing, can be an end
in itself, and the dictionary shows that that is a commonly accepted meaning
of the word.

> The issue is about the method. In this case, the method involves examining one
> item at a time and then placing it at a certain location depending on some
> criteria.
 
On the contrary, I'd say the IA Card Sort methodology is a process of
discovering just what those locations are. At the beginning of a card sort
there are *no* "certain locations". That's one of the main reasons for doing
the exercise.

>> Not quite sure what you mean by "the ultimate goal of sorting".
> 
> When order is not observed, there is no sorting.

If you mean sequence, then you are denying one dictionary meaning of the
word "sort". If instead you mean that after the process there is still only
chaos to be observed, then I'd agree. But you don't mean that, do you?

I do fully appreciate your point that grouping is often used as an
intermediary step towards sorting into sequences. Who hasn't sorted a deck
of cards by first grouping them into their suits, and then followed up with
the sequencing.

>> My dictionary has two definitions for "sorting". The first is
>> "placing things in categories according to shared attributes",
> 
> As I had pointed out earlier, grouping is a technique for sorting. To
> sort manually, people often use that technique. Thus, when they say
> they are sorting things, they often mean "placing things in categories
> according to shared attributes".

Substituting the definition in question for the word "sorting" you've said
"grouping is a technique for placing things in categories according to
shared attributes". Well, duh. Your statement only parses as non-empty if
you mean "sequencing" for "sorting". It still doesn't make sense though: My
dog has four legs. My other pet has four legs. Is my other pet a dog? If so,
why is it under water in a fish tank?

Grouping is also a technique for, well, grouping. No sequencing/sorting
intended, desired, or required. My kitchen has a drawer with sub-divisions.
In each sub-division there are knives, forks, and spoons. There is no
sequencing intended, required, or desired. This isn't simply because each
member of the categories are fungible either: I have soup spoons, dessert
spoons, and tea spoons; I have bread knives, steak knives, and regular
knives; I have fish forks and regular forks. No sequencing, only grouping.

...

BTW: I'm now very curious to know what techniques you would use for
discovery of categories. (Assuming that you're not allergic to the concept
of abstracting knowledge into sets or categories).

e.