Measurement
What are the four types of measurement scales?
Nominal, Ordinal, Interval, and Ratio
What is test reliability a measure of?
Consistency
What 2 things does Validity tell us?
What construct does the test measure, and how well does it measure the construct?
What are the 3 measures of Central Tendency?
Mean, Median, and Mode
What is the difference between a Criterion Referenced and a Norm Referenced test?
Norm Referenced tests compare scores to scores of similar individuals while Criterion Referenced tests compare scores to a set standard/criterion/mastery component.
What defines a Nominal scale of measurement?
A named category with no meaningful order
According to Classical Test Theory, every score is a combination of a true score and what?
Error or Measurement Error
Why can't a measure be considered inherently valid?
Tests are validated for specific purposes: the test must be valid for use in a specific context
What does a standard deviation tell us?
The distance of a score from the mean/how much a score varies from the mean
Difference from Variance: standard deviation tells us this information in the same units as our original scores!
What is a stratified random sample?
A sample selected based on specific characteristics/demographics
This is different from a cluster sample because cluster samples select existing clusters for convenience (i.e., specific school districts, specific classes, specific locations etc) rather than basing on individual characteristics (i.e., race, ethnicity, diagnosis, etc).
What defines a Ratio scale of measurement?
Units with a definite order, equal intervals, and a meaningful zero point
With this type of reliability, individuals are assessed with one instrument and then assessed immediately after with an equivalent version of the assessment
Alternate Forms or Parallel Forms
What is the difference between convergent validity and discriminant validity?
Convergent validity looks at constructs that should be associated, and discriminant validity looks at constructs that should be unrelated
What are the mean and standard deviation for a Z Score?
For Z scores, the Mean is always 0 and the Standard Deviation is always 1
What is the range of a correlation coefficient?
-1 to +1
Give an example of an Ordinal measurement
Any unit with a definite order but no equal interval (place in a race, Likert scale, etc)
What is Split-Half Reliability?
One instrument is split in half and the results of each half are correlated
Measures internal consistency
Give an example of Construct Irrelevance
Any extraneous factor that could affect a score (i.e., readability, social desirability, method of test administration, etc)
What are the mean and standard deviation of a T Score?
For T scores, the mean is always 50 and the standard deviation is always 10
If a test requires some level of subjective professional judgment, what type of reliability should be included in the manual?
Interrater Reliability
What is the difference between an Ordinal scale and an Interval scale?
Ordinal scales have a recognizable order but no regular intervals. Interval scales have both a recognizable order and regular intervals.
Name 2 considerations to keep in mind when using a Test-Retest method for determining reliability
No more than six months should pass between administrations of the test
The trait being measured needs to be stable over time
No impact of practice effects
No differential in learning between administrations
What are 5 types of evidence counselors must consider to determine a test's validity?
Item content, relation to other constructs, response processes, internal structure, and consequences of testing
What does a percentile score of 67 mean?
67% of scores on this assessment fall below this score
If an instrument measures consistently but does not measure what it was designed to measure, the instrument is...
Reliable, but not valid