Reliability
The degree to which a measurement is consistent
- the measure yields the same score for a given attribute everytime
Inter-rater reliability
the extent to which two clinicians agree on the diagnosis of an individual
- the similarity of two raters score
REASONS for if its Low:
- inadequate diagnosis manual
- diff. interview styles, interpretations, or important placed on the symptoms
Test-retest reliability
the ability of a measurement to yield the same score on two difference ocassions for the same individual
PROBLEMS:
- practice and familiarity with the test
Alternate form reliability
the extent to which two measures of the same attribute are correlated/yield the same results
- used to cirumvent the problems with test-retest
- create a second test to measure the same construct with slighly different questions - but measures the same thing
- highly correlated with the first test
Internal Consistency
Different parts of a test are correlated with eachother/produce the same results
- split half reliability
- alpha coefficient
Predictive Validity
the ability of a test to predict the course of an individiuals development
- crucial in diagnostic system - predict the progression of the disorder
Concurrent Validity
a test yields the same results as another measure of the same thoughts, feelings or behaviours
(two measures measuring the same thing ->highly correleated)
Validity
the extent to which the measurement measures what it purports to
- the ability of the diagnositc criteria to predict a mental disorder
Face Validity
Test includes characteristics that resemble the construct being measured
test APPEARS to be measuring the construct
Content Validity
Test includes a sample of behaviours that make up the construct/ are associated with the construct being measure
ex. test of depression - low energy, loss of interest, etc
Criterion Validity
Test yields a high score for people known to have that ability
Construct Validity
The test is measuring what it is supposed to be measuring and not something else
Validity
The degree to which a test measures what it is designed to measure
