|
Evidence based on test content
|
|
Logical analysis
|
Experts agreed the items were representative of the literature review and the target construct
|
|
Evidence based on internal structure
|
|
Item fit
|
Acceptable
|
|
Dimensionality
|
Three dimensions (personal conflict, family conflict, community conflict) supported
|
|
Precision
|
Acceptable
|
|
Reliability
|
Acceptable
|
|
Measurement invariance
|
Present. Pre-post comparisons of SECM measures would be appropriate
|
|
Evidence based on relationships with other variables
|
|
Convergence with measure of similar construct
|
Family conflict was significantly correlated with a modified IOS item. Evidence of convergence needed for personal and community conflict
|
|
Evidence based on response processes
|
|
Respondent cognition related to test
|
Partially addressed (i.e., “community”). Evidence is also needed to evaluate respondents’ interpretations of items not studied (e.g., using methods such as “think-aloud” interviews)
|
|
Validity generalization
|
|
Validity studies in different geographic, institutional, and demographic contexts
|
Not addressed. Evidence needed to evaluate if the inferences made from the instrument generalize to other contexts
|
|
Evidence of consequences
|
|
Outcomes of instrument use
|
Not addressed. Consequences (e.g., positive, negative) of instrument implementation for respondents and associated educational programs and systems
|
