Unit 4 Flashcards
Response
movement cycle
An instance or occurrence of a target behavior that
has a distinct beginning and end.
Permanent products
“A change in the environment produced by a
behavior that lasts long enough for measurement to
take place.”
Continuous (direct) response measures
“Directly measure a dimensional quantity of
behavior.”
Discontinuous (indirect) response measures
Measure a dimensionless quantity of behavior.
Types of continuous response measures
- Event recording (frequency recording)
- Latency recording
- Duration recording (per occurrence or total)
- IRT recording
Event recording (frequency recording)
A continuous (direct) response measure which measures rate.
Latency recording
A continuous (direct) response measure which measures latency.
Duration recording
A continuous (direct) response measure which measures duration.
Types of duration recording
- Duration per occurrence
* Total duration
IRT recording
A continuous (direct) response measure which measures IRT.
Types of discontinuous response measures
- Percent occurrence (and % correct)
- Trials to criterion
- Discrete categorization (coding)
- Partial interval recording
- Whole interval recording
- Momentary time sampling
- PLACHECK
Percent occurrence
A discontinuous (indirect) response measure which
measures the percentage of times a target behavior
occurs when an opportunity is presented for its
occurrence.
Trials to criterion
A discontinuous (indirect) response measure which
measures the number of consecutive opportunities to
respond, required to achieve a performance standard.
Discrete categorization (coding)
A discontinuous (indirect) response measure in which
responses are classified into discrete categories (e.g.,
severity codes, duration codes, independence codes,
etc.), and a percentage of these response categories
is obtained.
Partial interval recording
A discontinuous (indirect) response measure which measures the percentage of intervals in which a target behavior occurred at any time during an interval.
Whole interval recording
A discontinuous (indirect) response measure which measures the percentage of intervals in which a target response occurred for the total duration of an interval.
Momentary time sampling
A discontinuous (indirect) response measure which measures the percentage of intervals in which a target response was occurring at the end of an interval.
PLACHECK (Planned Activity Check)
A discontinuous (indirect) response measure which
measures the percentage of individuals that were
engaged in a target response at the end of an interval
of time.
Considerations in selecting response
measures
• Which dimensions are you interested in?
• What are the logistics involved in collecting data
(e.g., where is data going to be collected, who is
going to collect data, etc.)
Inter-observer agreement (IOA)
“The coefficient of agreement between two or more
independent observers.”
Formulas to calculate inter-observer
agreement
- (Smaller/larger) x100
* (Agreements/agreements + disagreements) x100.
How to establish adequate inter-observer
agreement
- Good response definitions
- Train observers
- Test observer agreement
- Frequent observer evaluation
Property
“A fundamental quality of a phenomenon.”
Dimensional quantity
“A quantifiable aspect of a property.”
Fundamental properties of behavior
• Temporal locus • Temporal extent • Repeatability • The combination of temporal locus and temporal extent
Dimensional quantities of behavior
- Latency
- Duration
- Countability
- Inter-response time
- Rate
- Celeration (acceleration or deceleration)
Temporal locus
“A response occurs at a certain point in time in
relation to a preceding environmental event.”
Temporal Extent
“A response occupies time.”
Repeatability
“A response can reoccur.”
Latency
“The amount of time between a stimulus and a
response.”
Duration
“The amount of time between the beginning and the
end of the response cycle.”
Countability (frequency)
“The number of responses or the number of cycles
of the response class.”
Inter-response time (IRT)
“The time between two successive responses;
Usually, the time elapsed between the end of a
response cycle and the beginning of the next
response cycle.”
Rate
“The ratio of the number of responses over some
period of time.”
Celeration
Changes in rate over time. Usually, increases
(acceleration) or decreases (deceleration) in rate
over time.
Reliability
“The consistency of measurement.”
Validity
“The extent to which data obtained from
measurement are directly relevant to the target
behavior of interest and to the reason(s) for
measuring it.”
Accuracy
The correspondence between observed instances
and “true” values.
Graph
“Relatively simple visual formats for displaying data.”
Types of graphic displays
- Bar graph (histogram)
- Cumulative record
- Equal-interval line graph
- Semi-logarithmic graph
Bar graph (histogram)
• Used for summarizing and comparing data for
different groups of subjects and/or different
conditions.
• Used to compare sets of data which are not
related to one another by a common underlying
dimension by which the horizontal axis can be
scaled. That is, the data are discrete.
Cumulative record
“A type of graph on which the cumulative number of
responses emitted is represented on the vertical axis;
the steeper the slope of the data path, the greater
the response rate.”
Cumulative recorder
“A device that automatically draws cumulative
records (graphs) that show the rate of response in
real time; each time a response is emitted, a pen
moves upward across paper that continuously moves
at a constant speed.”
Equal-interval line graph
It is the most common graphic
format for displaying data in Applied Behavior
Analysis.”
Permits the display of level, trend and variability.
Interpretation is a function of the way data is displayed.
Semi-logarithmic graph
“A two-dimensional graph with a logarithmic scaled y
axis so that equal distances on the vertical axis
represent changes in behavior that are of equal
proportions.
Uses of equal-interval line graphs
- Track client’s progress
- Evaluate effectiveness of interventions
- Present results of functional analyses
Interpreting graphs
• Did a significant change in one of the dimensions
of the target behavior occur between one
phase/condition and the next?
• Was the change due to the manipulation? How
certain are we?
Considerations during visual analysis of data
- Number of data points
- Level and mean shift (in level)
- Variability
- Trend
Level
The value on the vertical axis scale around which a
set of behavioral measures converge (mean or
median).
Trend
“The overall direction taken by the data path.”
Variability
“The extent to which measures of behavior under
the same environmental conditions diverge from one
another.”
Scatter plot
A data collection form for problem behavior and the
intervals of time behaviors occurred. It provides a
graphic display of data in a grid format. It is used to
identify patterns of responding in natural settings.
Bounce
Variability
The Standard Celeration Chart
“A multiply-divide chart with six base- 10 (or x10,
÷10) cycles n the vertical axis that can accommodate
response rates as low as 1 per 24 hours (0.000695
per minute) to as high as 1,000 per minute. It enables
the standardized charting of celeration, a factor by
which rate of behavior multiplies or divides per unit
of time.”
Houghton and Lindsley (1970)
Precision Teaching
Instructional decision-making system predicated on
the position that:
• Learning is best measured as a change in response
rate
• Learning most often occurs through proportional
changes in behavior
• Past changes in performance can project future
learning.
True value
A measure obtained by procedures that are
independent of and different from procedures that
produced the data being evaluated and for which the
researcher has taken “special or extraordinary
precautions to insure that all possible sources of
error have been avoided or removed.”
Celeration line
- At least 6 data points
1. Count number of total days
2. Split in half
3. Split in half again
4. Mark middle data point on middle day in each half
5. Connect two pints in step 4.
Ignore days vs. no opportunity days
Ignored day = bx occurred, no data collection, connect on graph
No chance = no opportunity for bx, do not connect
Record floor
Longest duration a response can occur in an observation period
