Scale development and Evaluation Flashcards
Test construction Strategies (3)
(1) Correspondence Strategy
(2) Empirical Strategy
(3) Construct (Theoretical) Strategy
Correspondence Strategy
One-to-one correspondence between person’s verbal reports and internal states
Correspondence Strategy was dominant in the _______
early 20th century
Correspondence Strategy is ________ and _______
Intuitive; atheoretical
Assumptions in Correspondence Strategy (4)
(1) Each item corresponds to a specific construct
(2) Item has COMMON MEANING for all test-takers
(3) A test-taker is able to accurately assess the requested information
(4) A test-taker will honestly report requested information
Empirical Strategy def
Items selected on basis of relations to external criteria. (e.g., contrasted groups)
In the empirical strategy, the meaning of an item is NOT equal to _________. It is determined by ______
the verbal content of the item; groups who endorse the item.
=> Interpretation of scores is by ‘cookbook“: EMPIRICALLY KNOWN CORRELATES of high and low scores
Concerns in Empirical strategy (3)
(1) Unintended group differences
(2) Problem of generalization
(3) Item overlap
Factor Analysis is used in _____
Empirical strategy
Factor Analysis def
Statistical procedure that aims to determine the number of distinct factors underlying.
“Factors”, in Factor analysis means: ___
Hypothetical variable reflecting a latent construct
Construct (Theoretical) Strategy originated in the ________
1950s; prominent from the 1960s to the present.
Construct strategy emphasizes ____________
theoretical definitions of domains to be measured
Goal of construct strategy
Measure an attribute – a hypothetical construct
Assumptions in Construct strategy (3)
(1) A person possesses some degree of a construct (e.g., sociability)
(2) Nontest behaviours can be identified which are referents (indicator) for the construct
(3) Test responses are referents (indicator) for the construct.
What does the construct strategy for reliability and validity?
Incorporates statistical methods
How does the construct strategy evaluates the adequacy of a test?
Evaluates adequacy of test by how well test fits in with theoretical (nomologic) net for the construct
7 Typical Steps in Theoretical Scale Construction Approach
(1) Define construct: Consider the literature for definition + theoretical relations with other constructs
(2) Gather/Write items
(3) Evaluate content validity: expert judgment to know if items are relevant
(4) Pre-testing of items: Administer initial pool to a small sample and conduct cognitive interviews
(5) Item reduction: Consider endorsement rate for items.
(6) Factor analysis: Determine the optimal number of factors underlying item response patterns
(7) Scale evaluation: test of dimensionality, reliability and validity