2-8 tb Flashcards
construct validity
extent to which evidence can be provided that test measures a theoretical constraint
extent to which evidence can be provided that test measures a theoretical constraint
construct validity
construct
an attribute, trait, or characteristic that in itself is not directly observable but can only be inferred by looking at observable behaviors, which are hypothesized to indicate the presence of that construct.
ex: we never really observe the construct of aggression, but we do observe behaviors that would lead us to conclude that a person might be demonstrating the construct of aggression
an attribute, trait, or characteristic that in itself is not directly observable but can only be inferred by looking at observable behaviors, which are hypothesized to indicate the presence of that construct.
construct
abstract constructs
underlying attitudes or attributes that exist in our imaginations (intelligence, beauty, love)
underlying attitudes or attributes that exist in our imaginations (intelligence, beauty, love)
abstract constructs
concrete constructs
activities that are observable and measurable
activities that are observable and measurable
concrete constructs
What are the similarities and differences between abstract constructs and concrete constructs?
Sim: they both measure
Diff: abstract is more theoretical while concrete is observable
construct explication
defining or explaining a psychological construct
defining or explaining a psychological construct
construct explication
What is the three step process for constructor’s explication?
- identify the behaviors that relate to the construct
- identify other constructs that may be related to the construct being explained
- identify behaviors related to similar constructs, and determine whether these behavior are related to the ongoing construct
gathering evidence of construct validity
two types: theoretical evidence and psychometric evidence
this is an ongoing process, not an event
gathering theoretical evidence
establish nomological network: lis the associations the construct has with other constructs
propose experimental hypothesis: propose one or more hypotheses using the test as an instrument for measuring the construct
establish nomological network: lis the associations the construct has with other constructs
propose experimental hypothesis: propose one or more hypotheses using the test as an instrument for measuring the construct
gathering theoretical evidence
gathering psychometric evidence
convergent evidence of validity & discriminant evidence of validity
convergent evidence of validity & discriminant evidence of validity
gathering psychometric evidence
gathering psychometric evidence: Convergent Evidence
shows that constructs that theoretically should be related are related; evidence that the scores on a test correlate strongly with scores on other tests that measure the same constructs
methods used to gather: ????
test correlates consistent with theory
shows that constructs that theoretically should be related are related; evidence that the scores on a test correlate strongly with scores on other tests that measure the same constructs
gathering psychometric evidence: Convergent Evidence
gathering psychometric evidence: Discriminate Evidence
when the test scores do not correlate with unrelated constructs
methods used to gather: correlate test scores
when the test scores do not correlate with unrelated constructs
gathering psychometric evidence: Discriminate Evidence
multitrait-multimethod design (MTMM)
tests for convergence across different measures of the same thing and for divergence between measures of related but conceptually distinct things
researcher chooses two or more constructs that are unrelated and two or more types of tests to measure each of the constructs
tests for convergence across different measures of the same thing and for divergence between measures of related but conceptually distinct things
multitrait-multimethod design (MTMM)
factor
the underlying commonalities of tests or test questions that measure a construct
What a group of test items have in common that accounts for their correlations
the underlying commonalities of tests or test questions that measure a construct
What a group of test items have in common that accounts for their correlations
factor
factor analysis
an advanced statistical procedure based on the concept of correlation that helps investigator to explain why two tests are correlated
an advanced statistical procedure based on the concept of correlation that helps investigator to explain why two tests are correlated
factor analysis
Differences and similarities of CFA and EFA
Diff:
- CFA researcher specifies in advanced what the factor should look like
- CFA can specify in advance a large number of characteristics believed to be present in the data
- CFA can specify the amount and nature of any measurement error present i the data
Sim:
- both specify the number of factors they expect to be able to extract from data
- both explain correlations between items or tests
