[DDI-SRG] Variables, questions and qeustionnaire

Wendy Thomas wlt at pop.umn.edu
Thu Jan 8 11:25:46 EST 2009


It should probably reference the controlconstruct (if available) as well 
as question. Referencing the question indicates that the same question was 
asked (valuable for comparison between surveys). However, only the 
controlconstruct provides the universe information. Now...a reference to a 
control construct leads one to the question making the question reference 
redundant. But we don't always have the control construct. Also the 
universe is dependent upon the use of a control construct within a master 
series (inclusion of individual constrolconstructs by reference). We need 
to make sure all of this works from both directions (survey production and 
end user).

Wendy

On Thu, 8 Jan 2009, Pascal Heus wrote:

> All:
> Was thinking about the variable/questionnaire relationship this morning
> and we may have a problem in the schema as we can define a link between
> a question and a variable (using  Variable/QuestionReference) but I
> don't think we can related the variable to the survey instrument itself
> (like the ControlConstruct and the actual location in an Instrument).
> This is important when I serialize a file (actually collect and store
> data) as it is typically then tied to a specific Instrument
> (questionnaire). If I reuse a question in multiple ControlContruct and
> then insert that in Instrument(s) (one or more time), how do I know
> which actual question in my instrument the variable in my data file
> refers to? I don't think this is possible in 3.0.
>
> Use case: a simple household survey:
> - Define question Q1: "how old is [NAME]"
> - Define variable V1: age  that refers to Q1
> - Define two constructs that use Q1
>   - CC1 used in the demographics module, question is asked to head of
> household)
>   - CC2 , used in the children under 16 questionnaire (asked to the kids)
> - Define Instrument I1 that uses CC1 and CC2 (and/or make case more
> complex by creating more than one instrument)
> - When I actually produce data for that instrument where I reuse the age
> variable V1, it can come from two different places and I do not know
> which one
>
> (1) Is this really a problem (or did I miss somtehing)?
> (2) If yes, I think this could be resolved at the LogicalRecord or
> Variable (maybe by refining the Question reference or adding reference
> to the CC and/or Instrument) [need some thinking]
>
> Note that if I create two age variables, the problem remains as I don't
> know which CC / instrument it refers to
>
> *P
>
> _______________________________________________
> 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


More information about the DDI-SRG mailing list