accuracy of our measurements
consistency of our measurements
reliability
Degree to which what we are
measuring actually measures
what we want to measure!
Validity
how well a test is consistent within itself
internal reliability
a test that appears to be a good measure has good what?
Face validity
What is the variable(s) that is manipulated in an experiment?
Independent variable
A general tendency to respond in the same way,
irrespective of the question being asked.
What is Response Bias
Tests that measure factors such as:
Personality traits
Mental disorders
Attitudes
Intelligence
What is Psychometric Tests
stability of responses, and whether the same pattern of answers can be obtained through repeated
testing
external reliability
a test that identifies a performance outcome has what validity.
predictive validity
What is the outcome variable in an experiment?
Dependent variable
True or false: Reverse scoring items increases likelihood of response bias?
What is false
A test scale that ranges from strongly disagree to strongly agree?
What is a Likert scale.
two (or more) researchers observe and measure the same phenomena, and then correlate the degree of similarity in their respective ratings.
Inter-rater
performances that you capture in a laboratory be different from how they are in the ‘real’ world?
Ecological validity
In a normal distribution what percentage of the population falls within 2 standard (1.96 to be exact) deviations above and below the mean?
what is 95%
What is the measure that usually has responses in opposites?
What is semantic differential scales
Compares an individuals score to scores of other similar individuals
Normative values
administer the test at one point in time, and again at some pre-determined later time.
test-retest reliability
A measure that correlates to a particular construct has the same outcome of another test?
Convergent/discriminant validity
a p value of 0.15 is what?
not significant
measure can either be too high or low
overall they tend to cancel each other out is a form of what error?
Random error
Measurement theory states that
"true score" = test score (+error)
Essentially an average of all possible
split-half calculations
Cronbach's alpha
Does the measure distinguish between
respondents in relation to relevant criteria?
Criterion Validity
arbitrary predetermined threshold most commonly used in psychology?
alpha level .05