Chapter 5- Improving and Assessing the Quality of Behavioral Measurement Flashcards
Validity
Data is directly relevant to the phenomenon measured and to the reason(s) for measuring it.
Accuracy
Extent to which the observed values matches the true value of the event.
Measurement bias
Nonrandom measurement error, likely to be in one direction.
Reliability
The consistency of measurement, specifically, the extent to which repeated measurement of the same event yields the same value.
Indirect measurement
Occurs when the observer measures a proxy, or stand-in, for the actual behavior of interest. Secondhand information that requires the observer to make inferences.
Direct measurement
Occurs when the behavior measured is exactly the same as the behavior that is the focus of the behavior change program.
Continuous measurement
Measurement that detects all instances of the target behavior during the observation period.
Discontinuous measurement
Any form of measurement in which some instances of the response class of interest may not be detected.
Observer drift
The unintended changes in the way data are collected may produce measurement error.
Naive observer
A trained observer who is unaware of the study’s purpose of experimental effect during a given observation period.
Calibration
Comparing the data produced by a measurement system to a known standard of true value and, if necessary, adjusting the measurement system so the data it produces match the known standard.
Total cont IOA
A percentage of agreement between the total number of responses recorded by two observers.
(smaller count/ larger count) x100
Mean count-per-interval IOA
(int 1 IOA + int 2 IOA + int N IOA/ n intervals)
Exact count-per-interval IOA
(number of intervals of 100% IOA/ n intervals) x100
Trial-by-trial IOA
(number of trials agreement/ total number of trials) x100
Total duration IOA
(shorter duration/ longer duration) x100
Mean duration-per-occurrence IOA
(dur IOA R1 + dur IOA R2 + dur IOA Rn/ n responses with dur IOA) x100
Interval-by-interval IOA
(number of intervals agreed/ total number of intervals) x100
Scored-interval IOA
Only the intervals in which either or both observers recorded the occurrence of the target behavior
(agreed occurrence/ total trials with occurrences) x100
Unscored-interval IOA
Only the intervals in which either or both observer recorded the nonoccurrence of the target behavior
(agreed nonoccurrence/total trials with nonoccurrences) x100