exam 1 study pt 4

face validity

does the test look like it measures what its supposed to

what are the types of evidence for validity

3 cs

• content validity

• criterion validity

• construct validity

what is criterion validity

comapres new test to “gold standard”

• there are 2 types

• concurrent and predictive

describe concurrent and predictive types of criterion validity

concurrent- both new and gold standard test given at same time (if highly correlated, good criterion validity)

predictive- can test predict future outcomes/reference standard

Ex:fall risk test predicts future falls

what is construct validity , what are the two types

does the test measure theoretical concept (aka the construct)

convergent and divergent

what is the convergent and divergent types of construct validity

convergent - should correlate with similar measures ( depression test 1 should equal depression test 2.0)

divergent- should not correlate with unrelated measures ( depression test should not correlate to grip strength test)

what are constructs that cannot be measured

• depression

• anxiety

• quality of life

• etc

what is content validity

the extent to which a test, survey, or assessment comprehensively covers all aspects of the subject it aims to measure

-

-

• Definition: Checks if a test covers all parts of the idea or trait it is meant to measure.

• Method: Experts review the test questions to see if anything important is missing or off-topic.

what is criterion referenced

compared to fixed benchmark or standard

what is norm referenced

compared to other ppl, how one person compares to another

what is a minimally clinincally important difference (MCID)

smallest change that is meaningful to the patient

what is change score

change score= follow up score- baseline score

aka treatment effectiveness, disease progression,

ceiling and floor effect

little room for. imp or decline

factors affecting validity of change

level of measurment

reliabiliyt

stability

baseline scores

potential sources of measurement error

• person takign measurments

• measuring instrument

• variability in characteristic being measure

what is relative reliability

ICC, KAPPA, measurments, unitless

measures how consistently subjects maintain their position or rank order in a group, even if their exact scores change

what is absolute reliability

measured in SEM, original units

how much error is in the source

what to what value is poor reliability, moderate reliability, and good reliability

below .50- poor reliability

0.50-0.75- moderate reliabilty

.75 and higher - good reliability

what are the types of reliability

test - retest

rater

alternate forms

internal consistency

what is test re-test reliability

same test to same ppl at 2 different times

what is rater reliabilty and types

intra rater- 1 rater measures same thing multiple times

inter rater- 2 examiners score the same performance

what is alternate forms reliability

two diff versions of the same test given to the same ppl

what is internal consistency reliability

how well multiple items within a single test or survey measure the same concept.

measured with Cronbach's Alpha (α): aka the reliability rating from 0-1