Levels of Measurement
Attributes are names or categories
Nominal
Clear and detailed
Specific
The degree to which an instrument produces consistent results.
Reliability
How accurate an instrument is at measuring what it is trying to measure.
Validity
Intervention only
B design
Attributes can be ordered
Ordinal
Quantifiable; ask yourself, how will you measure the results?
Measurable
When the same instrument is given to a group of people at 2 different times.
Test-retest
Least scientific type; researchers simply look at an instrument and give their opinion on its validity.
Face validity
Baseline and intervention
AB design
Equal intervals; distance is meaningful
Interval
Realistic given resources and abilities
Attainable/Achievable
When a group of people complete 2 similar versions of an instrument.
Parallel-forms
When we are able to generalize about our construct of interest because we are accurately measuring that construct.
Construct validity
Basic withdrawal
ABA design
Equal intervals with absolute zero
Ratio
Goal should be linked to your role in the work setting or your personal life (if it's a personal goal)
Relevant
When different instrument items that are trying to measure the same construct are compared to see how they produce similar results.
Internal consistency
When the results from the instrument accurately relates/predicts some kind of external variable.
Criterion-related validity
Set a deadline to evaluate progress
Time-limited
When the decisions from different raters are compared to each other to see how consistent the raters' decisions are.
Inter-rater reliability
How well an instrument covers the range of meanings included within a concept that is being measured.
Content validity