I've seen a few (unanswered) requests in this thread for relevant research
on this topic -- is there none? Phrases "I like to..." and "I can imagine a
user who..." quickly lead to disaster, forming the basis for why usability
testing is so powerful.
Or is the relevant research too hard to find because our search/browse
interfaces are so awful? :-)
On Fri, Mar 13, 2009 at 12:05 PM, Karen Coyle <lists_at_kcoyle.net> wrote:
> Owen, yes, I think this is worth exploring. One of the bright lights of
> early librarianship (no time to dig it up, maybe later) talked about "known
> order," mainly referring, of course, to alphabetical order. Back to the
> phone book: if your language has an alphabetical order (and not all do),
> then taking words/names and placing them in that order is helpful because
> users know the alphabet. If instead you ordered the phone book by how tall
> people are, it wouldn't be very helpful because the order wouldn't match the
> user's knowledge. The other thing about this kind of search is that the user
> probably has to have a pretty good idea of what he's looking for -- a
> particular place in the known order, e.g. someone's last name. This is
> really a 'find' not a 'browse' in my mind. No one expects to go from a-z in
> the phone book, although they may wish to in a shorter list (your journal
> titles example). Funny, though, that we don't expect items on a restaurant
> menu to be in alphabetical order, yet we are usually able to find what we
> are looking for ("do they have a tuna sandwich"?). I think that *quantity*
> of entries is a big factor in terms of how much order is needed.
>
> It seems to be inherent in a thesaurus that the users are not expected to
> already know the order of the entries, but that the thesaurus actually
> guides the user. The user isn't doing a mere lookup, but is following a
> structure. The user possibly expects to be directed to other concepts from
> his entry term and the thesaurus must be able to guide the user from
> whatever starting point the user chooses. So if someone goes into the
> thesaurus with the term "sociology" the thesaurus will provide a conceptual
> context and some directions the user can go in. Thesauri tend to be
> hierarchical, but I think we could do ones that are not so using more
> relationships than just BT, NT, RT.
>
> Some folks have experimented with providing LC classification has a kind of
> subject browse. I don't know how useful that has turned out to be for users.
> It is multiple hierarchies and not perfectly hierarchical, but it has
> conceptual structure. But LCSH isn't LCC. The main thing is that LCSH is
> neither a 'known order' list (because users don't know the entry terms) nor
> is it a thesaurus, because it has so little conceptual structure. So trying
> this thought experiment with LCSH in mind gets one pretty twisted.
>
> kc
>
>
> Stephens, Owen wrote:
>
>> Thanks Ross
>>
>> I completely agree! I'm playing Devil's advocate a bit here - because I
>> think the issues are worth exploring.
>> I don't think, as you say, "browse indexes" per se are the problem - but
>> thesaurus browse presents special problems in terms of User Interface design
>> (I don't feel for example that Bernhard has agreed on this yet, and I'm not
>> clear if this is because he disagrees or because I haven't been clear enough
>> - if he, or others, disagree then it would be interesting to understand why)
>>
>> If we agree on this (or at least that thesaurus browse offers specific
>> challenges for User Interface) what the best ways of presenting the power of
>> the thesauri to the user is so they can exploit it to its fullest potential?
>>
>> Finally - the big question - how do I find a good plumber in the greater
>> Manhattan area?
>>
>> Owen
>>
>>
>>
>>
>>> The difference between the journal A-Z list (or a phone book) and a
>>> thesaurus browse is that vocabularies generally offer an additional
>>> axis -- up and down.
>>>
>>> Couple this to the fact the thesaurus in particular (in this case,
>>> LCSH) is granular as hell and sometimes "obtusely phrased" with
>>> crossreferences all over the place and I think you have a very alien
>>> interface to most users.
>>>
>>> Then add in the fact that most subject browses don't actually show the
>>> resources they're referring to (i.e. the bib records) -- although
>>> obviously this is technically feasible.
>>>
>>> I think an easy distinction is how much easier it is (to take the
>>> phone book analogy a little further) to find something in the white
>>> pages (known item/A-Z) vs. the yellow pages (hierarchical subject then
>>> alpha browse). The larger the metropolitan area, the lower the
>>> usability of the yellow pages.
>>>
>>> -Ross.
>>>
>>> On Fri, Mar 13, 2009 at 10:55 AM, Stephens, Owen
>>> <o.stephens_at_imperial.ac.uk> wrote:
>>>
>>>
>>>> Karen,
>>>>
>>>> I generally agree with all this. I tend to think exploring from a
>>>>
>>>>
>>> 'known item' or 'known set of items' is what people do in the physical
>>> environment (identifying a collection, shelf, or other physical
>>> starting point to browse). However, I also believe that in certain
>>> situations the concept of a browse of headings works for the user and
>>> they are happy with it. For example (and I can't explain it!)
>>> persistently users like to have an A-Z list of e-journals - every time
>>> I suggest they can use the catalogue, I run up against the fact that
>>> the users really seem to prefer an A-Z list - I don't know why.
>>>
>>>
>>>> What I'm trying to explore is where the problems with browsing the
>>>>
>>>>
>>> 'headings' list lies. What we've seen from at least a couple of people
>>> is an argument that presented properly the LCSH (or indeed Name
>>> authority files) are a powerful (and essential?) tool for all users.
>>> We've also seen Bernhard's straw poll that librarians like browse
>>> indexes (but still not, I don't think, nailed down what exactly a
>>> browse index is, and whether we are all talking about the same thing)
>>>
>>>
>>>> You say "users aren't looking for headings" - I tend to agree - I'd
>>>>
>>>>
>>> be interested to hear any counter arguments? (the only counter argument
>>> I can think of is that if users look for headings they do more
>>> effective searching because buy not using the headings they risk
>>> missing items - but personally although I might accept that this is
>>> currently true, it isn't where I think we should be aiming).
>>>
>>>
>>>> Personally my current feeling is that these authority files and
>>>>
>>>>
>>> structured thesauri can work hard for the user, but that the user does
>>> not need to be exposed to them for this to happen. I am, however, open
>>> to persuasion :) Having worked in Medical libraries a while ago, my
>>> instinct is that for searching specialist literature there is more to
>>> be said for explicitly interacting with the taxonomy - but does this
>>> just contradict my starting position?
>>>
>>>
>>>> Owen
>>>>
>>>> Owen Stephens
>>>> Assistant Director: eStrategy and Information Resources
>>>> Central Library
>>>> Imperial College London
>>>> South Kensington Campus
>>>> London
>>>> SW7 2AZ
>>>>
>>>> t: +44 (0)20 7594 8829
>>>> e: o.stephens_at_imperial.ac.uk
>>>>
>>>>
>>>>
>>>>> -----Original Message-----
>>>>> From: Next generation catalogs for libraries
>>>>> [mailto:NGC4LIB_at_LISTSERV.ND.EDU] On Behalf Of Karen Coyle
>>>>> Sent: 13 March 2009 14:06
>>>>> To: NGC4LIB_at_LISTSERV.ND.EDU
>>>>> Subject: Re: [NGC4LIB] Browse functionality (was Whose elephant is
>>>>>
>>>>>
>>>> it,
>>>
>>>
>>>> anyway? (the OLE project))
>>>>>
>>>>> Stephens, Owen wrote:
>>>>>
>>>>>
>>>>>> However, browsing a structured thesaurus/authority file/taxonomy -
>>>>>>
>>>>>>
>>>>> call it what you like - is less common or straightforward. What
>>>>>
>>>>>
>>>> isn't
>>>
>>>
>>>> clear to me is whether this is about implementation or about the
>>>>>
>>>>>
>>>> nature
>>>
>>>
>>>> of structured taxonomies (my guess is that it is a combination). In
>>>>> general I don't think library users need teaching how to use a
>>>>> telephone book - but when the telephone book starts to say things
>>>>>
>>>>>
>>>> like
>>>
>>>
>>>> "See, See Also, and Narrower Term References:
>>>>>
>>>>>
>>>>>> * Broader Terms not currently available" when you look for Leo
>>>>>>
>>>>>>
>>>>> Tolstoy's number it starts to get complicated :)
>>>>> Owen, as I said in a post somewhere along the line, we should
>>>>> think
>>>>> about what it is we are asking users to browse (or offering to them
>>>>>
>>>>>
>>>> to
>>>
>>>
>>>> browse): catalog headings. I suspect that people are fine browsing
>>>>> something that they already understand (phone book), but my
>>>>>
>>>>>
>>>> experience
>>>
>>>
>>>> is that users don't understand the dis-embodied headings list, since
>>>>> they have seen the catalog entry as the 'unit' they are seeking.
>>>>> Essentially, users aren't looking for headings, they are looking for
>>>>> books/movies/music. I think only librarians and 'super-users' look
>>>>> explicitly for headings. It probably made more sense to users of the
>>>>> card catalog because they saw the headings at the top of the cards,
>>>>>
>>>>>
>>>> but
>>>
>>>
>>>> in our case users doing a browse in the online catalog did not
>>>>> understand what they were looking at. Add to that the fact that
>>>>>
>>>>>
>>>> library
>>>
>>>
>>>> cataloging headings can be rather artificial and obtuse, and that
>>>>> browse
>>>>> lists in large catalogs are unwieldy ("Hamlet. 1603" "Hamlet. 1604"
>>>>> "Hamlet.... " for pages) and users just give up.
>>>>>
>>>>> Here are some thinks I think users would be happy to browse:
>>>>> - all the books by an author
>>>>> - all the books in a 'series' (e.g. Harry Potter)
>>>>> - everything new put on the shelf in the last week
>>>>>
>>>>> We treat these as 'retrievals' but this is what users come to the
>>>>> library to browse, and I suspect that's what browse means to them.
>>>>>
>>>>> kc
>>>>>
>>>>>
>>>>>
>>>>>> We are talking of "index browsing", not "browsing" in all its
>>>>>>>
>>>>>>>
>>>>>> broad,
>>>
>>>
>>>> general, everyday usage. Think of this like browsing the index of
>>>>>>> a book, an alphanumerical arrangement. Only that in a database
>>>>>>>
>>>>>>>
>>>>>> index
>>>
>>>
>>>> you don't physically flip pages but click links to get the
>>>>>>>
>>>>>>>
>>>>>> results
>>>
>>>
>>>> behind the index entries. Or, for that matter, to be taken
>>>>>>>
>>>>>>>
>>>>>> forthwith
>>>
>>>
>>>> to
>>>>>
>>>>>
>>>>>> the synonym or alternate form of name or whatever. The point is
>>>>>>>
>>>>>>>
>>>>>> that
>>>
>>>
>>>> the index gives you serendipitous context from which to go on
>>>>>>> and explore terms you didn't happen to think of but find useful.
>>>>>>> Something to which keyword searching, for all its merits, just
>>>>>>>
>>>>>>>
>>>>>> cannot
>>>>>
>>>>>
>>>>>> be
>>>>>>> equivalent - for it cannot show you what you missed. Good index
>>>>>>> browsing
>>>>>>> offers help and insight in an unobtrusive way - once the user
>>>>>>> understands what an index actually is. If the latter is
>>>>>>>
>>>>>>>
>>>>>> inachievable,
>>>>>
>>>>>
>>>>>> then of course that concept is doomed, for better or worse.
>>>>>>>
>>>>>>>
>>>>>> (Consult
>>>
>>>
>>>> Jims latest posting do decide which.)
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> OK - I really feel we need to get more specific terminology here -
>>>>>>
>>>>>>
>>>>> for me "index browsing" suggests the ability to browse a list of
>>>>>
>>>>>
>>>> terms
>>>
>>>
>>>> pulled from records you have in the database. This could be a
>>>>>
>>>>>
>>>> keyword
>>>
>>>
>>>> index - but I'm not sure browsing a keyword index is really very
>>>>> sensible or useful.
>>>>>
>>>>>
>>>>>> As far as I can see you are not talking about this, but talking
>>>>>>
>>>>>>
>>>>> about
>>>
>>>
>>>> taxonomy browsing? We have already established I think that an
>>>>>
>>>>>
>>>> actual
>>>
>>>
>>>> alphabetical list of LCSH is not what you are looking for, but
>>>>>
>>>>>
>>>> rather
>>>
>>>
>>>> some way of exploring LCSH in a way that hides some of the
>>>>>
>>>>>
>>>> complexity
>>>
>>>
>>>> until the time it is useful to the user?
>>>>>
>>>>>
>>>>>> I think the idea that keyword searching cannot show you what you
>>>>>>
>>>>>>
>>>>> missed needs exploring - as instinctively I believe this is a
>>>>>
>>>>>
>>>> function
>>>
>>>
>>>> of the User experience not of the approach they have taken. This is
>>>>>
>>>>>
>>>> one
>>>
>>>
>>>> of the things that the 'facets' uncovered by the type of NGCs we are
>>>>> currently seeing - however the question of how this helps the user
>>>>>
>>>>>
>>>> is a
>>>
>>>
>>>> slightly different one.
>>>>>
>>>>>
>>>>>> To take for example the NLA VuFind implementation - since I've got
>>>>>>
>>>>>>
>>>>> it
>>>
>>>
>>>> to hand :)
>>>>>
>>>>>
>>>>>> Let's say I'm interested in the book "Daredevils of the skies" by
>>>>>>
>>>>>>
>>>>> Norman Ellison and similar literature.
>>>>>
>>>>>
>>>>>> If I do a keyword search for 'Daredevils' then one of the facets I
>>>>>>
>>>>>>
>>>>> see is the subject heading "Air pilots -- Australia -- Biography --
>>>>> Juvenile literature". At the moment this implementation only allows
>>>>>
>>>>>
>>>> me
>>>
>>>
>>>> to narrow my search to this subject heading - but there is
>>>>>
>>>>>
>>>> absolutely
>>>
>>>
>>>> no reason why there shouldn't also be an option to 'broaden' my
>>>>>
>>>>>
>>>> search
>>>
>>>
>>>> to include all items with this heading - this is an user
>>>>> experience/interface design choice - nothing to do with where I
>>>>>
>>>>>
>>>> started
>>>
>>>
>>>> my search.
>>>>>
>>>>>
>>>>>> However, even without this 'Broaden search to this subject
>>>>>>
>>>>>>
>>>>> heading'
>>>
>>>
>>>> option, the NLA implementation does do several other things that
>>>>>
>>>>>
>>>> allow
>>>
>>>
>>>> me to explore the NLA collection serendipitously - the 'Finding
>>>>>
>>>>>
>>>> aids'
>>>
>>>
>>>> points up that the Papers of Norman Ellison are available *and* if I
>>>>> click on the book 'Daredevils of the skies' and look at 'Similar
>>>>>
>>>>>
>>>> Items'
>>>
>>>
>>>> I see that I can view a collection of Norman Ellison's slides - this
>>>>>
>>>>>
>>>> is
>>>
>>>
>>>> clearly showing me stuff that I would have otherwise 'missed'.
>>>>>
>>>>>
>>>>>> I'm not arguing that these latter points of exploration should
>>>>>>
>>>>>>
>>>>> replace structured taxonomies - I guess I'm undecided on the merits
>>>>>
>>>>>
>>>> of
>>>
>>>
>>>> this - but I am convinced that if our structured headings are going
>>>>>
>>>>>
>>>> to
>>>
>>>
>>>> be successfully exploited we need to understand what value they add,
>>>>> exactly how they add it, and then design a user experience that
>>>>> exploits this. I haven't seen anything that achieves the last of
>>>>>
>>>>>
>>>> these
>>>
>>>
>>>> yet...
>>>>>
>>>>>
>>>>>> Owen
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>> --
>>>>> -----------------------------------
>>>>> Karen Coyle / Digital Library Consultant
>>>>> kcoyle@kcoyle.net http://www.kcoyle.net
>>>>> ph.: 510-540-7596 skype: kcoylenet
>>>>> fx.: 510-848-3913
>>>>> mo.: 510-435-8234
>>>>> ------------------------------------
>>>>>
>>>>>
>>>>
>>
>>
>>
>
>
> --
> -----------------------------------
> Karen Coyle / Digital Library Consultant
> kcoyle@kcoyle.net http://www.kcoyle.net
> ph.: 510-540-7596 skype: kcoylenet
> fx.: 510-848-3913
> mo.: 510-435-8234
> ------------------------------------
>
--
Bill Dueber
Library Systems Programmer
University of Michigan Library
Received on Sat Mar 14 2009 - 11:16:18 EDT