topic 8 Flashcards
even when we observe the right set of units
observation might be imperfect, random errors or systematic errors
systematic errors
measuring the wrong construct
reliability
consistency of a measurable instrument, same of similar results when used repitedly under the same conditions
how do we divide relibabilty
1 stability when we use same operation. several times, test retest reliability, inter/rater reliability (asking several observers to rate the same object)
2 CONSISTENCY use and compare similar operations, items are designed to measure the same thing
validity
accuracy and truthfulness of the measurement instrument, valid instrument should capture the correct underlying concept
internal validity
true for context of same study
external validity
findings are applicable to others contexts
accurate observation
precise, reliable, valid
approaches to measure validity
face validity, content, construct criterion/related
can validity be measured
the value is unknown so it cannot be not observed directly
face validity
appers to measure what intended, APPEARS, based on purely subjective assessment
Content validity
covers all of the theoretical dimensions, covers all dimension DEFINE CONCEPT, EXPERT REVIEW, PILOT TESTING, REVISE AND REFINE
construct validity
goes deeper, if it measures the theoretical construct , accurately assesses the measurement tools
criterion related validity
correlation with standards, how well correlates with an external criterion,
criterion related validity two types
predictive how well the instrument predicts the outcome, concurrent validity - how well correlates with other established measures at the same time
