[DDI-SRG] Terms: life cycle, XML Schema/schemes

[Wendy Thomas]

There are three other recurring themes in terms of questions that we 
should probably have an agreed and consistant response to. The first is 
common when doing short presentations because we do not go into extensive 
content detail (its too much and there's no time):

"It's a 'black box'"
"It's way too complicated to use"
"How do you get the information from the researcher?"
"What's the point without tools?"
"Where are the tools?"

Here are the responses I've been using, but as I said, we need to craft 
responses that are clear, easy to remember, and can be used consistantly. 
These are not necessarily those answers, just what I've been responding.

"It's a 'black box'"
In the sense of this presentation, yes it is, because there wasn't time to 
go into detail. However, the contents of this particular black box are 
well structured and predictable. The components capture pieces of 
information and relationships in a consistant way that allows programmers 
to build tools to capture information, run search systems, and develop 
complex software to explore and anylize the data being described.

"It's way too complicated to use"
You are rarely going to be looking at DDI content directly in the XML 
structure. You will normally be using tools for input or exploration and 
the content will be displayed through an interface. The underlying 
structure was designed to facilitate reuse of the metadata. Viewers such 
as XSLT let you arrange the content and look at it in a variety of 
formats. What's going on behind what you see in a word processor is also 
complicated, you just don't think about it because you aren't look at it 
directly.

"How do you get the information from the researcher?"
This is not a new problem. It's been with us regardless of the format of 
the metadata we are collecting. Earlier versions of DDI helped by 
identifying specific pieces of metadata that were important to capture. It 
was clearer than a general description. However, to get this inforamtion 
from researchers we need to make the process of capturing the metadata 
something that facilitates their work. Large data collection agencies only 
became really interested when they learned that DDI 3.0 could assist them 
in maintaining consistancy within the organization (through registries) 
and that the content was structured in a way that they could not only 
reuse the metadata rather than reenter it, but use it to inform and drive 
their production processes. For individual researchers we need to identify 
the direct advantages to their research process and build tools that will 
allow them to take advantage of this.

"What's the point without tools?"
Its a bit like the chicken and the egg, you need to have a structure for 
the content before you can start building tools.

"Where are the tools?"
The DDI Alliance recognized early on that without tools there would not be 
a lot of DDI 3.0 users. The candidate release and review involved the 
development of some tools and common code that would be needed for tools 
development. The early tools focus on moving metadata into and out of 
statistical packages (SPSS, SAS, etc.). The DDI Foundation Tools program 
was designed to coordingate and support early open source tools 
development so that we could pool our talents and get the initial tools 
out as soon as possible. A DDI editor is a priority project and parts are 
currently under development. [Reference tools update information, current 
status and estimated delivery dates if known]


I think it would be more effective having a set of bullet points for each 
of these common questions rather than a script. Mostly we want to be sure 
people are getting the same 2 or 3 basic points from whoever is being 
asked within the DDI Alliance.

Wendy


On Wed, 2 Jul 2008, Joachim Wackerow wrote:

> Looking back at several DDI workshops and presentations for mixed
> audiences I have the impression that two often used core terms should be
> clarified in expression, and used in a constant and clear manner.
>
> Suggestion for discussion (this is the outcome of talking with Wendy
> about that):
>
> "life cycle" should be "data/metadata life cycle"
>
> Reasoning: Researcher are more focused on people not metadata. So
> sometimes people think at life cycle of people (not data/metadata) like
> life history. The documentation is concerned about the life cycle of
> data. Reuse of documentation in the DDI sense is concerned about the
> life cycle of metadata. Regarding life cycle, data and metadata can
> probably not be easily differentiated in different terms. Therefore
> combination of data/metadata. Sometimes the term "life cycle of survey"
> is used. This should be really only used when dealing with survey data.
> DDI can be used for a broader range of data. Therefore the term data is
> here used (not survey) in a general sense.
> Note: sometimes people say that "life cycle" is not a good metaphor
> because in a biological sense there is death at the end but no death of
> data. I think there is no better metaphor for that. A possible response
> would be: life cycle on the level of the species not the individual.
>
>
> XML Schemas versus schemes (like variable scheme)
> "XML Schemas" should be always "XML Schemas" not just "Schemas"
> schemes should be "DDI schemes"
>
> Reasoning: These similar sounding terms can cause some confusion.  "XML
> Schema" is a well defined term from W3C. It is a technical means to
> define a XML structure. It seems to be quite common to use it in this
> writing. Schemes in DDI are basically lists of items like variable
> schemes (list of variables), questions schemes etc. The term scheme in
> DDI is used for exactly this purpose. This must be explained in
> documentations and teaching materials. The fixed term "DDI schemes"
> should be used always when mentioning these (not just schemes).
>
>
> When we can agree on new terms they should be used in a constant manner
> throughout all documentation and teaching material.
>
> Probably other terms are also candidates for clear and consequent wording.
>
> Any thoughts or other ideas?
>
> Achim
> _______________________________________________
> DDI-SRG mailing list
> DDI-SRG at icpsr.umich.edu
> http://www.icpsr.umich.edu/mailman/listinfo/ddi-srg
>

Wendy L. Thomas                          Phone: +1 612.624.4389
Data Access Core Director		 Fax:   +1 612.626.8375
Minnesota Population Center              Email: wlt at pop.umn.edu
University of Minnesota
50 Willey Hall
225 19th Avenue South
Minneapolis, MN 55455