GF 3 Flashcards
(31 cards)
statements that attempt to explain and predict behavior
Theories
hypothetical attributes such as motivation or attraction
Constructs
measures the construct by measuring an observable, external behavior
Operational definition
prediction about the relationship between 2 or more constructs
Conceptual hypothesis
specific prediction about the relationship between 2 precisely defined constructs
Operational hypothesis
the degree to which the measurement procedure accurately measures the variable it’s supposed to measure
Validity
What determines the quality of a measurement procedure?
Validity and reliability
the consistency of a measurement
Reliability
when, on the surface, a measurement seems to be valid
Face validity
when a new measurement is accepted as valid because it reflects the results of a similar “gold standard”, well-established measurement
Concurrent validity
when a measurement actually predicts the behavior caused by the construct
Predictive validity
Negative relationship
still shows a consistent concurrent relationship, near -1
when the measurement items accurately reflect the content of the construct
Content validity
when every measurement procedure behaves in the same way as everything we know about the construct
Construct validity
What are 2 types of construct validity?
Divergent and convergent
when we see that 2 measurements measure the same construct
Convergent validity
when we see that 1 measurement measures 2 different things
Divergent validity
What are ways to improve validity?
Test re-test, multiple observers, physical measures, population estimates
What are types of errors?
observer error, test characteristics, participant changes, environmental changes
What are the 2 qualitative scales of measurement?
Nominal (can’t be put in order) and ordinal (can be put in order)
What are the 2 quantitative scales of measurement?
Interval and ratio
reliability measure where 2 different versions of an instrument are used for test re-test
Parallel-forms reliability
the degree of agreement between 2 separate observers simultaneously recording experiments
Inter-rater reliability
when researchers split a test in half and compute separate scores for each half to see if they have similar results
Split-half reliability