describing missing values for a rdfs:DataType (numreric/text domains)

[borsna]

If there is a numeric response domain we need to be able to describe missing values.
Any advice how to do this would be welcome.

Take into account that code/ categories missing are defined on disco:Concept isValid moving the definition of missing to representation means removing isValid on the Concept.

[LarryHoyle]

One thought:
keep in mind that some software (like SPSS) labels "in band" values as
missing - values that otherwise would be valid. Other software has one or
more system missing values - like R. Still others have multiple user
defined out of band missing values, like SAS and Stata, with missing values
.a through .z.

On Fri, Oct 25, 2013 at 3:43 AM, Olof Olsson notifications@github.comwrote:

If there is a numeric response domain we need to be able to describe
missing values.
Any advice how to do this would be welcome.

Take into account that code/ categories missing are defined on
disco:Concept isValid moving the definition of missing to representation
means removing isValid on the Concept.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/130
.

[kulnor]

I would in general you won't see this often but but agree that SPSS has
this option build in. You could also have a case where you can define a
unbounded conditional range like >9 is missing.

Note that a similar issue exists with code lists. For example, SAS
allows you to associate a band with a single category (like 10-99 =
Foo). This cannot be handle by DDI (and most other packages). The option
to code all values in range as many single values is not practical (as I
could define 9-99999 which could generate thousands of entries). But
this is not a new problem.

best
*P

On 10/25/13 7:28 AM, LarryHoyle wrote:

One thought:
keep in mind that some software (like SPSS) labels "in band" values as
missing - values that otherwise would be valid. Other software has one or
more system missing values - like R. Still others have multiple user
defined out of band missing values, like SAS and Stata, with missing
values
.a through .z.

On Fri, Oct 25, 2013 at 3:43 AM, Olof Olsson
notifications@github.comwrote:

If there is a numeric response domain we need to be able to describe
missing values.
Any advice how to do this would be welcome.

Take into account that code/ categories missing are defined on
disco:Concept isValid moving the definition of missing to
representation
means removing isValid on the Concept.

—
Reply to this email directly or view it on
GitHubhttps://github.com//issues/130
.

—
Reply to this email directly or view it on GitHub
#130 (comment).

[boschthomas]

@JoachimWackerow Achim, do you have a solution solving this problem Olof came up with?

[JoachimWackerow]

See diagram Figure 32 skos:Concept and skos:ConceptScheme, file concept.png.

Disco uses skos:Concept as representation of the DDI items concept, and code and category as well. This is the background that isValid is an additional property of skos:Concept for the use by code and category.

The code can also be used for a number of a numeric response domain. isValid would express if this code is valid or not.

Is something really missing?