[DDI-users] Using CategoryScheme, CodeScheme, and QuestionScheme and Ontology concept

Jeremy Iverson jeremy at algenta.com
Mon Jan 17 10:26:11 EST 2011


I would do this as follows. You can find a corresponding DDI 3.1 file 
attached. Please feel free to get in touch if you would like further 
details.

1. Create a ConceptScheme with the terms of your ontology.

2. Create a single CategoryScheme to represent possible genders.

3. Create a separate CategoryScheme to represent missing values.

4. Create two CodeSchemes: one mapping 1=Male, 2=Female, etc; the other 
mapping 0=Male, 1=Female, etc.

5. Create two QuestionSchemes representing the questions as held by each 
Center. The first QuestionScheme will hold question q_1, which will 
reference the first CodeScheme as its ResponseDomain. The second 
QuestionScheme will hold q_10, which will reference the second 
CodeScheme. Both questions will reference the same terms from your 
ontology (Concepts): gender, male, and female.

6. Create two Instruments to represent the questionnaires as 
administered by each Center. The first will simply contain q_1, and the 
second will contain q_10.


Cheers,

Jeremy
-- 
Jeremy Iverson

Algenta Technologies
1428 Washington Ave S Ste 203
Minneapolis, MN 55454

+1 (888) 264-0711 ext 4
jeremy at algenta.com

http://www.colectica.com/software/reader - the free DDI 3 viewer




On 1/16/2011 9:28 PM, My Coyne wrote:
> Hi All,
>
> I'm a DDI newbie and am in search for a way to express the following
> scenario using DDI. I will take a simplest case of gender.
>
> a. Gender may have values of male, female, both, unknown, and missing
> values. The missing values may also have a lot more values, for
> simplicity, says 'don't feel like answering'.
>
> b. In the project, I would also like to use ontology. Gender, male, and
> female all can be expressed as concept in an ontology.
>
> c. The project has different study centers; each study center has its
> own way to label male and female and not necessary same codes. For example,
> c.1. Study center UC1 in their instrument I_1, question q_1
> q_1 Gender: Male (recorded in the database as 1)
> Female (recorded in the database as 2)
> c.2 Study center SC2 in their instrument I_2, question q_10
> q_10 Gender: Male (recorded in the database as 0)
> Female (recorded in the database as 1)
>
> I'm investigating the use of DDI for data archiving and achieving the
> following objectives:
>
> 1. Set up the metadata services for consistency going forward in the future
> 2. Still allowing for legacy data as being done at Study center SC1 and SC2
> 3. Setting a data framework to connect to an ontology
>
> I plan to use DDI in the following manner, please comment if it is
> correct or not. I appreciate any help. Thank you in advance.
>
> a. CategoryScheme to define each "genderMale", "genderFemale", "genderBoth"
> b. Use CategoryScheme to define gender that would make reference to
> "genderMale", "genderFemale", and "genderBoth"
> c. Using CodeScheme and codeItem to code 'Male" that refers to
> categorySchem genderMale; "Female" that refers to genderFemale, etc.
> d. Define QuestionScheme "GenderQuestion"
>
> But I fail to solve the problem that Q_1 (in instrument I_1) will have
> Male as 1 and Q_2 (in instrument I_2) will have a code of 0-zero. I
> defined (a) and (b) as gender is a concept in ontology, so are "male"
> and "female".
>
> Thank you for any help.
>
>
> My Coyne
> mcoyne at boninc.com
>
>
>
> _______________________________________________
> DDI-users mailing list
> DDI-users at icpsr.umich.edu
> http://www.icpsr.umich.edu/mailman/listinfo/ddi-users
-------------- next part --------------
A non-text attachment was scrubbed...
Name: genderSample.xml
Type: text/xml
Size: 22848 bytes
Desc: not available
Url : http://www.icpsr.umich.edu/pipermail/ddi-users/attachments/20110117/b88a598c/attachment-0001.xml 


More information about the DDI-users mailing list