[DDI-CDG] [Fwd: some comments]

Tue Sep 6 07:46:49 EDT 2005

Dear all!

over the past days I have received three reactions to the draft of our 
proposal. Reto has sent his around already. This one below is from Wendy 
and there is a third one from Meinhard, which I attach to this mail (I 
hope that's o.k. Wendy?). It is the proposal with some remarks.

To structure today's discussion I would suggest to have each respondent 
explain a little about his comments. This will probably cover a lot of 
the problems. After that I would like to see whether there are any 
points missing and how we can formulate stuff.
1. Replies (in alphabetical order)
- Meinhard
- Reto
- Wendy
2. Open items in the paper

I will collect your suggestions, work them into the draft and send it 
around asap. I would also like to propose another call early next week 
to go through those changes and see that we get the proposal done. This 
does not mean that we have to put a halt to ideas and thoughts, all the 
contrary. The finished model by the SRD will be fed to the Alliance for 
decision and we will see, if it fits all ours needs (save the whales, 
ban nuclear arms, you know ;-). Just kidding!

Talk to you soon,
Oliver.

-------- Original Message --------
Subject: some comments
Date: Mon, 5 Sep 2005 09:21:14 -0500 (CDT)
From: Wendy Thomas <wlt at pop.umn.edu>
To: Oliver Watteler <watteler at za.uni-koeln.de>

Hi Oliver,

You'll be happy to know you have me "laboring" on Labor Day :-) I've been
reading through your document. You did a great job of filtering and
sorting out all of these discussions. A couple of comments:

1) I've attached the current drafts of the documentation suggestion for
time and geography. These are the reusable classes available in all
modules. Take a look and a) check to see that containers for descrete
information needed by the comparative group are covered, and b) what types
of links (pointers) need to be in specific modules are needed to make use
of these. For example you note some specific comparative time/geography
issues...what types of special information or identifer is needed in the
main module that can be linked to the geography or time classes.

My impression is that the comparative group is going to be primarily
concerned (in terms of elements/attributes) with making sure the
variables, geogrpahy, time, and instrument are completely and accurately
defined. The issue they becomes how comparision aspects are expressed.
Does this become a separate module? I think, in terms of "life cycle"
and wanting to make as few changes to "persistent" sections, that a
"comparison" module (or probably the "grouping" module) is most
appropriate. It would need to point to questions, variables, and
methodology information in other modules and then contain "comparative"
information. What is this comparative information?

Link between 2 or more like items (variable a to variable b)
Type of relationship:
    equivilent
    parent | child | sibling | component part | ??

Use of the relationship
    used for joining (time, geography, responseUnit to responseUnit)
    used in composed variable
    used in a recode

It seems from both conference call discussions, email (both on and off the
list) and outside discussions, that we will end up using a combined
hierarchical and pointer system to reflect multiple hierarchies. While
there is some desire to make a choice between the two (the choice would
ahve to be pointer since we don't deal with clean hierarchies), but it is
probably more pragmatic to do a combined.

A question for you...since I don't want to throw a wrench in the
works...would it be feasible to create a section in the "grouping" module
that desribes the multiple hierarchical relationships and their
relationships in a consistant way (rather than specifically nesting
concept or other modules). Just a thought. I guess in effect that makes
the whole thing a pointer system, but it does allow for clear mapping of
simple hierarchies in a single place rather than having to traverse the
document as a whole.

And finally, identification of what must be machine actionable and what
can/should be human readable. Sometimes both are required, sometimes one
will do.

Wendy

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

-- 

GESIS - Zentralarchiv für Empirische Sozialforschung (ZA)
   Oliver Watteler, M.A.

   Dokumentation und Archivierung
  /Documentation and Archiving

   Bachemer Str.40
   50931 Köln
   Tel.:++49-221-47694-76
   FAX :++49-221-47694-44
   Web :http://www.gesis.org/za

-------------- next part --------------
A non-text attachment was scrubbed...
Name: Geography_09_02_2005.xls
Type: application/octet-stream
Size: 20992 bytes
Desc: not available
Url : http://www.icpsr.umich.edu/pipermail/ddi-cdg/attachments/20050906/eda8156e/Geography_09_02_2005-0001.obj
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Time_08_26_2005.xls
Type: application/octet-stream
Size: 21504 bytes
Desc: not available
Url : http://www.icpsr.umich.edu/pipermail/ddi-cdg/attachments/20050906/eda8156e/Time_08_26_2005-0001.obj
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Proposal_draft version 1_31082005_mm.doc
Type: application/msword
Size: 59904 bytes
Desc: not available
Url : http://www.icpsr.umich.edu/pipermail/ddi-cdg/attachments/20050906/eda8156e/Proposal_draftversion1_31082005_mm-0001.doc