Question: Is there a simple way to select subsets of cases from a data file for Facets analysis? Specifically, if I had a dataset in which I embedded a group ID in one of the facet labels or in the examinee ID, could I select only the cases where the group ID is a single number?

For example, suppose I created a data set like this:

1ID1,…

1ID2,…

2ID3,…

2ID4,…

Is there a simple way to tell Facets only to use the cases where the first column of the first variable is 1?

 

Answer: DVALUES= can help you.

Define a dummy facet of "Group ID"

Facets = 4  ; examinee, item, rater, Group ID

Models =

?, ?, ?, 1, R4   ; selects only group 1 for analysis

*

Labels =

1, Examinee

1= 1ID1,…

2= 1ID2,…

3 =2ID3,…

4 = 2ID4,…

....

*

2, item

...

*

3, rater

....

*

4, Group ID, A  or D ; dummy facet used only for selection (can also be used for interactions, etc.)

1 = Group 1, 0

2 = Group  2, 0

3 =  Group 3, 0

4 =  Group 4, 0

*

Dvalues =

4, 1, 1    ; the element identifier for facet 4 is in facet 1,  the first character of the element label

*

Data =

examinee, item, rater, observation  ; do not specify an element number for facet 4 here

.....