[DDI-SRG] DataKind

[Sanda Ionescu]

Yes, I understand, except it's very hard to generalize starting from a
Census example, as this is a "special" case.
Yet, it is true that if the DataKind is "census/enumeration" then the
product may be either micro- or aggregate. That was my point too.
Microdata or aggregate data describe the product.
But in non-census studies we can have multiple dataKinds - for instance,
both survey and administrative records, or experiments, etc.
and only one type of data product - numeric, and microdata.
Usually, but not necessarily, the different dataKinds would be stored in
different physical files. 
So, that's my two cents. I do not have a definitive solution.
But it's definitely a topic for further discussion. This is an issue
that needs to be clarified.

Sanda.

Sanda Ionescu
ICPSR
University of Michigan
P.O. Box 1248
Ann Arbor, MI 48106

Phone, Fax: 734-615-7890 
-----Original Message-----
From: Wendy Thomas [mailto:wlt at pop.umn.edu] 
Sent: Thursday, January 03, 2008 1:25 PM
To: Sanda Ionescu
Cc: ddi-srg at icpsr.umich.edu; Mary Vardigan
Subject: RE: DataKind

Sanda,

re: the last statement...just my cryptic shorthand :)

In 2.1 if you had the 1990 PUMS 5% and the 1990 STF3 you'd have two DDI
instances each with one data collection source 1990 US Census and one
logical product (either PUMS 5% or STF3). In DDI 3.0 you'd have one DDI
instance covering the 1990 US Census containing multiple data products
(PUMS 5%, STF3, etc.) Each product would have a kind of data (microdata,
aggregate data, but come from the same data source census/enumeration
and survey.

Wendy


On Thu, 3 Jan 2008, Sanda Ionescu wrote:

> Wendy, Mary, and others.
>
> Looking more closely at the enumeration below, I think there is some
> inconsistency, in that most concepts indicate how the data were
> collected (by means of a survey, or an experiment, or observations,
> etc.) while "aggregate data" does refer to the data product, and
should
> be a counterpart to "microdata" - possibly within a different
> classification.
> Maybe we should re-analyze this list within the Controlled
Vocabularies
> group and see what makes sense. But we definitely need a DataKind
entry,
> I think.
>
> I may be thick, but I'm not sure exactly what you mean by "in  3.0
there
> can be a n..1 (product to source) relationship". What is the "product"
> you refer to? a data file? can you give an example?
>
> Sanda.
>
>
> Sanda Ionescu
> ICPSR
> University of Michigan
> P.O. Box 1248
> Ann Arbor, MI 48106
>
> Phone, Fax: 734-615-7890
> -----Original Message-----
> From: Wendy Thomas [mailto:wlt at pop.umn.edu]
> Sent: Thursday, January 03, 2008 12:05 PM
> To: ddi-srg at icpsr.umich.edu
> Cc: Mary Vardigan; Sanda Ionescu
> Subject: DataKind
>
> I am adding the bug on missing DataKind and have a question. In DDI
2.0
> there was a different StudyUnit for each logical product (e.g.
> microdata, aggregate, etc from the 1990 Census). Does DataKind refer
to
> the data SOURCE or the data PRODUCT?
>
> <!-- 2.2.3.10 Kind of Data.
> -->
> <!-- The type of data included in the file: survey data,
> census/enumeration
>       data, aggregate data, clinical data, event/transaction data,
> program
>       source code, machine-readable text, administrative records data,
>       experimental data, psychological test, textual data, coded
> textual,
>       coded documents, time budget diaries, observation data/ratings,
>       process-produced data, etc.
> -->
>
> This is a subelement of Summary Description
>
> 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
>

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