Validity and Reliability

The degree an instrument measures what it is supposed to measure
-content
-criterion-related
-construct

Established when the researchers know that the instrument measures the concept intended to be measured.
- give the instrument to a panel of experts, who judge the instrument by rating each item being meausred

-established by letting other colleagues or subjects examine the instrument and indicate whether it appears to measure the concept
-less desirable b/c uses an intuitive approach

The degree to which the observed score and the true score are related.
-tested for: concurrent validity and predictive validity

-researchers simultaneously administer two different instruments measuring the same concept.
-use correlations to compare scores from the 2 insturments
(high correlations=agreement, and low correlations= the instruments are measuring different concepts.)

Refers to whether a current score is correlated with a score obtained in the future
-give an instrument today and again in 2 weeks---should see less of a correlation

-focuses on theory
-theoretical concepts that are tested empirically
-to what extent does the instrument measure the theoretical concept or trait?

Researchers use theories to make perdictions about the concept being measured
-data are gathered, and a determination is made as to whether the findings support the hypothesis

-researchers use two or more instruments to measure the same theoretical component
-focuses on how the observed score compares to the theory
-similar to concurrent testing

Comparing scores from two or more instruments that measure different theoretical constructs
-ex: depression and happiness

-when convergent and divergent testing are combined
-helpful in reducing systematic error
-has 3 instruments

-administer instruments to indiv. known to be either high or low on the characteristic being measured
-expect there to be significant difference between them

-used to identify questions that group around different factors
-a statistical approach
-items that group together have high correlations
-questions that do not fit into a group are altered or eliminated.

-when researchers obatin consistent measurements over time.
-the instrument can be reliable w/o being valid
-ex: scale measuring your weight but not your anxiety
expressed in correlation coefficients (0-1)
0= no correlation, 1= perfect correlation
0.70 or greater is considered acceptable