Quiz - T-test, ANOVA, Validity, Reliability, Objectivity Flashcards
Population
Entire group of individuals/observations in question
Parameters
Measure of INTEREST in the population
-Ex: BF % in CSUF students (populations)
Sample
Subgroup of population
-Important if randomly selected
Statistics
Measure of interest in sample
-Ex: Avg or mean BF % (measure) is the statistics
Independent Variable (x)
- The Cause
- What you manipulate
Dependent Variable (y)
-Outcome/consequence
Validity
The degree when the test MEASURES what is SUPPOSED TO MEASURE
Most important characteristic of a test
Logical Validity
Exs: We want to determine who jumps highest so we do vertical jump test
-Obvious a jump test so it is LOGICAL a vertical jump test can determine that
Content Validity
- Taking an exam w/materials that don’t cover info
- Missing materials on exam
- Imbalanced: disproportionate distribution
- Not only written, can be physical
Concurrent Validity
Degree to which scores on a test CORRELATE w/scores on an ACCEPTED STANDARD (CRITERION MEASURE)
- Interested in this when using a test in place of another test bc it is more accurate
- Exs: Underwater test more accurate than skinfold but both are concurrent validity
Predictive Validity
Degree to which one can predict future performance on a second measure
-Exs: Heart disease (criterion: heart disease developed later in life). Present diet, exercise behaviors, blood pressure, family history is checked
Construct Validity
- Degree to which a test measure a part of a whole skill or an abstract trait
- Ex: measures something that isn’t directly observable (emotions, personality, attitude)
Reliability
The degree of CONSISTENCY w/which a test measures what it measures
-Just bc a test is reliable, doesn’t mean its VALID
Stability Reliability
- Degree to which an individual’s score are UNCHANGED DAY-TO-DAY
- Should never be -1.0 ever for reliability
Internal-Consistency Reliability
- Degree to which an individual’s scores are UNCHAGED WITHIN A DAY
- At least 2 trials a day
