Measurements Flashcards
3 Dimensional Quantities
RTT
Repeatability
Temporal Extent
Temporal Locus
Repeatability
When behavior can be COUNTED
3 Types of Repeatability
CRC
Count
Rate
Celeration
Count
Add up behaviors
Does not give enough information to make decisions regarding a clients behavior
Rate
Frequency Add up behaviors or items over time count/time Most popular Must be standard Used to record free operants Should not be used for discrete trials
Celeration
Count per unit of time/time
Frequency/time
How rates of response change over time
Celeration trend line through the graphed data shows direction and degree of trend
Temporal Extent
Duration of behavior can be measured
Duration- amount of time in which a behavior occurs
Total- Cumulative amount of time in a total session
Duration-per-occurence- duration of target behavior
Temporal Locus
Measuring the time at which behavior occurs
Locus=point in time
2 Types of Temporal Locus Measures
Response Latency
Interresponse Time
Response Latency
Time between onset of a stimulus and initiation of response
Interresponse Time
IRT Amount of time that elapses between 2 consecutive instances of a response class When time between responses is important
2 Derivative Measures
Percentage
Trials- To- Criterion
Percentage
Ratio formed by combining the same dimensional quantities such as count
Expresses proportional quantity of an event
Recommended minimum observation= 30
Trials-To-Criterion
A measure of number of response opportunities needed to achieve a predetermined level of performance criteria
Trial depends on the nature of target behavior and desired performance level
2 Definitional Measures
Topography
Magnitude
Topography
Form or shape of the response
Magnitude
Force
Intensity
Severity of behavior
Continuous Measure Procedures
Measurement conducted in a manner such that all instance of response class are detected
Useful for free operant, discrete beginning and end
Not useful for high rate behaviors, discrete trials, opportunity based
Discontinuous Measure Procedures
Measurement conducted in a manner such that some instances of response class of interest may not be detected
Useful for high rate behaviors
Not useful for free operants
3 Procedures for Measuring Behavior
ETT
Event Recording
Timing
Time Sampling
Event Recording
Methods to record the number of times a response occurs
Not useful for very high rates of behavior
Timing
Used to measure
Duration, Response Latency, IRT
Time Sampling
Variety of methods for recording behavior during intervals or at specific moments in time
Time sampling methods gives approx.
DIvide observation into equal intervals
3 Forms of Times Sampling
WPM
Whole Interval Recording
Partial Interval Recording
Momentary Time Sampling
Whole interval Recording
Divide observation period into a series of brief intervals
At the end of each interval record whether the target behavior occurred throughout the interval
Always percentage
Underestimates rate of behavior
Partial Interval Recording
Divide observation period in intervals and record whether occured at any time during the interval
Used to rep. proportion
Always percentage
Momentary Time Sampling
Divide observation period in intervals
At the end of each interval record whether the target behavior occurred at the end of that time interval only
Always percentage
Planned Activity Check
For groups a variation of Momentary Time Sampling
Permanent Product
Measuring behavior after it occurred by measuring effects behavior produced on environment
Ex post facto measurement
Event recording, timing and time sampling can be applied to measurement by permanent product
When To Use Permanent Product?
Are you interested in only the outcome of the behaviors?
Is real time measurement needed?
Can the target behavior be measured by pp?
3 Indicators of Trustworthy Measurement
VAR
Validity
Accuracy
Reliability
Validity
Directly measuring target behavior
Measuring dimension relevant to the question
Ensuring data representative of behaviors during times that are relevant
Threats to Validity
Indirect Measurement
Measuring Wrong Dimension
Measurement Artifacts
Measuring Artifacts
Discontinuous Measurement
Poorly Schedules Measurement Periods
Insensitive and Limiting Measurement Scales
Accuracy
Observed value matches the true value
If not Valid, Accuracy is moot
Reliability
Procedures yield the same value when brought into REPEATED contact with the same state of nature
Repeatability
Threats to Reliability and Accuracy
Human Error
Poorly Designed measurement systems
Inadequate observer training
Expectations about what the data should look like
Observer Drift
When observers unknowingly alter the way they measure Bx
Measurement Bias
Nonrandom measurement error
Overestimating or underestimating the true value
Naive Observers minimizes measurement bias
Interobserver Agreement
Refers to the degree to which 2 or more independent observers report the same values after measuring the same events
AKA of Interobserver Agreement
IOA
4 Event Recording IOA’s
Total Count
Mean Count-per-Interval
Exact Count- per Interval
Trial by Trial
Total Count IOA
Simplest method
% of agreement between the total number of responses recorded by 2 observers
Overestimates the extent of actual agreement
smaller #/ Larger # X 100
Mean Count Per- Interval IOA
Dividing observation period into a series of small counting times
Int 1 IOA + Int 2 IOA+ Int N IOA/ Total # Intervals
Exact Count Per Interval IOA
of Intervals of 100 IOA agreement/ total # of Intervals X 100
Trial by Trial IOA
of Trials of Agreement/ Total # of Trials X 100
2 Timing/Duration IOA’s
Total Duration
Mean Duration Per Occurence
Total Duration IOA
Shorter/ Longer X 100
Mean Duration Per Occurrence IOA
Duration IOA B1+ Duration IOA B2+ Duration IOA Bn/ n Bx with Duration IOA
3 Time Sampling/ Interval Recording IOA’s
Interval by Interval IOA
Scored Interval IOA
Interval by Interval IOA
of Intervals Both Recorders are in Agreement/ Total # Intervals X 100
Scored Interval IOA
Intervals Both Recorders Recorded Occurrences/# of Intervals at least one recorder recorded occurring X 100
Unscored Interval IOA
of Intervals Both Recorders Recorded Non- Occurence/ # of Intervals at least one recorded Non- Occurrence X 10
