1

Cooper Summary
Part 1 Introduction and Basic Concepts

Definition and Characteristics of Applied Behavior Analysis

Science: Basic Characteristics and a Definition

Different types of scientific investigations yield knowledge that enables the description,
prediction, and/or control of the phenomena studied.
Descriptive studies yield a collection of facts about the observed events that can be quantified,
classified, and examined for possible relations with other known facts.
Knowledge gained from a study that finds the systematic covariation between two events—
termed a correlation—can be used to predict the probability that one event will occur based on
the occurrence of the other event.
Results of experiments that show that specific manipulations of one event (the independent
variable) produce a reliable change in another event (the dependent variable), and that the change
in the dependent variable was unlikely the result of extraneous factors (confounding variables)—
a finding known as a functional relation—can be used to control the phenomena under
investigation.
The behavior of scientists in all fields is characterized by a common set of assumptions and
attitudes:
 Determinism—the assumption that the universe is a lawful and orderly place in which
phenomena occur as a result of other events.
 Empiricism—the objective observation of the phenomena of interest.
 Experimentation—the controlled comparison of some measure of the phenomenon of
interest (the dependent variable) under two or more different conditions in which only
one factor at a time (the independent variable) differs from one condition to another.
 Replication—repeating experiments (and independent variable conditions within
experiments) to determine the reliability and usefulness of findings.
 Parsimony—simple, logical explanations must be ruled out, experimentally or
conceptually, before more complex or abstract explanations are considered.
 Philosophic Doubt—continually questioning the truthfulness and validity of all scientific
theory and knowledge.
 2

A Brief History of Behavior Analysis

Behavior analysis consists of three major branches: behaviorism, the experimental analysis of
behavior (EAB), and applied behavior analysis (ABA).
Watson espoused an early form of behaviorism known as stimulus–response (S–R) psychology,
which did not account for behavior without obvious antecedent causes.
Skinner founded the experimental analysis of behavior (EAB), a natural science approach for
discovering orderly and reliable relations between behavior and the environmental variables of
which it is a function.
EAB is characterized by these methodological features:
 Rate of response is the most common dependent variable.
 Repeated or continuous measurement is made of carefully defined response classes.
 Within-subject experimental comparisons are used instead of designs comparing the
behavior of experimental and control groups.
 The visual analysis of graphed data is preferred over statistical inference.
 A description of functional relations is valued over formal theory testing.
Through thousands of laboratory experiments, Skinner and his colleagues and students
discovered and verified the basic principles of operant behavior that provide the empirical
foundation for behavior analysis today.
Skinner wrote extensively about a philosophy for a science of behavior he called radical
behaviorism.
Radical Behaviorism attempts to explain all behavior, including private events such as thinking
and feeling.
Methodological Behaviorism is a philosophical position that considers behavioral events that
cannot be publicly observed to be outside the realm of the science.
Mentalism is an approach to understanding behavior that assumes that a mental, or “inner,”
dimension exists that differs from a behavioral dimension and that phenomena in this dimension
either directly cause or at least mediate some forms of behavior; it relies on hypothetical
constructs and explanatory fictions.
Pragmatism is the philosophical position that the truth or value of a scientific statement is
determined by the extent to which it promotes effective action.
The first published report of the application of operant conditioning with a human subject was a
study by Fuller (1949), in which an arm-raising response was conditioned in an adolescent with
profound disabilities.
 3

The formal beginnings of applied behavior analysis can be traced to 1959 and the publication of
Ayllon and Michael’s article “The Psychiatric Nurse as a Behavioral Engineer.” Contemporary
applied behavior analysis (ABA) began in 1968 with the publication of the first issue of the
Journal of Applied Behavior Analysis (JABA).
Characteristics of Applied Behavior Analysis
Baer, Wolf, and Risley (1968) stated that a research study or behavior change program should
meet seven defining dimensions to be considered applied behavior analysis:
 Applied—investigates socially significant behaviors with immediate importance to the
subject(s).
 Behavioral—entails precise measurement of the actual behavior in need of improvement
and documents that it was the subject’s behavior that changed.
 Analytic—demonstrates experimental control over the occurrence and nonoccurrence of
the behavior—that is, if a functional relation is demonstrated.
 Technological—the written description of all procedures used in the study is sufficiently
complete and detailed to enable others to replicate it.
 Conceptually Systematic—behavior change interventions are derived from basic
principles of behavior.
 Effective—improves behavior sufficiently to produce practical results for the
participant/client.
 Generality—produces behavior changes that last over time, appear in other
environments, and/or spread to other behaviors.
ABA offers society an approach toward solving many of its problems that is accountable, public,
doable, empowering, and optimistic.
A Definition of Applied Behavior Analysis
Applied behavior analysis is the science in which tactics derived from the principles of behavior
are applied systematically to improve socially significant behavior and experimentation is used
to identify the variables responsible for behavior change.
Behavior analysts work in one or more of four interrelated domains:
 Behaviorism (theoretical and philosophical issues),
 The Experimental Analysis Of Behavior (basic research)
 Applied Behavior Analysis (applied research)
 Professional Practice (providing behavior analytic services to consumers).
Translational Research bridges basic and applied research and informs both domains.
ABA’s natural science approach to discovering environmental variables that reliably influence
socially significant behavior and developing a technology to take practical advantage of those
discoveries offers humankind its best hope for solving many of its problems.
 4

Applied behavior analysis research and practice have improved human performance and the
quality of participants’ lives across a wide range of areas, but no problem has been solved
completely, and many important problems, challenges, and opportunities remain.

Basic Concepts and Principles

Behavior

Behavior is the activity of living organisms. Technically, behavior is “that portion of an

organism’s interaction with its environment that involves movement of some part of the
organism” (Johnston & Pennypacker, 2009, p. 31).
The term behavior is usually used in reference to a larger set or class of responses that share
certain topographical dimensions or functions.
Response refers to a specific instance of behavior. Response topography refers to the physical
shape or form of behavior.
A response class is a group of responses of varying topography, all of which produce the same
effect on the environment.
Repertoire can refer to all of the behaviors a person can do or to a set of behaviors relevant to a
particular setting or task.
Environment
Environment is the physical setting and circumstances in which the organism or referenced part
of the organism exists.
Stimulus is “an energy change that affects an organism through its receptor cells” (Michael,
2004, p. 7).
The environment influences behavior primarily by stimulus change, not static stimulus
conditions.
Stimulus events can be described formally (by their physical features), temporally (by when they
occur), and functionally (by their effects on behavior).
A stimulus class is a group of stimuli that share specified common elements along formal,
temporal, and/or functional dimensions.
Antecedent conditions or stimulus changes exist or occur prior to the behavior of interest.
Consequences are stimulus changes that follow a behavior of interest.
Stimulus changes can have one or both of two basic effects on behavior: (a) an immediate but
temporary effect of increasing or decreasing the current occurrences of the behavior and/or (b) a
 5

delayed but relatively permanent effect in terms of the future occurrences of that type of behavior
in the future.
Respondent Behavior
Respondent behavior is elicited by antecedent stimuli.
A reflex is a stimulus–response relation consisting of an antecedent stimulus and the respondent
behavior it elicits (e.g., bright light–pupil contraction).
All healthy members of a given species are born with the same repertoire of unconditioned
reflexes.
An unconditioned stimulus (e.g., food) and the respondent behavior it elicits (e.g., salivation) is
called an unconditioned reflex.
Conditioned reflexes are the product of respondent conditioning: a stimulus–stimulus pairing
procedure in which a neutral stimulus is presented with an unconditioned stimulus until the
neutral stimulus becomes a conditioned stimulus that elicits the conditioned response.
Pairing a neutral stimulus with a conditioned stimulus can also produce a conditioned reflex—a
process called higher-order (or secondary) respondent conditioning. Respondent extinction
occurs when a conditioned stimulus is presented repeatedly without the unconditioned stimulus
until the conditioned stimulus no longer elicits the conditioned response.
Operant Behavior
Operant behavior is selected by its consequences. Unlike respondent behavior, whose topography
and basic functions are predetermined, operant behavior can take a virtually unlimited range of
forms.
Selection of behavior by consequences operates during the lifetime of the individual organism
(ontogeny) and is a conceptual parallel to Darwin’s natural selection in the evolutionary history
of a species (phylogeny).
Operant conditioning, which encompasses reinforcement and punishment, refers to the process
and selective effects of consequences on behavior:
 Consequences affect future behavior.
 Consequences select response classes, not individual responses.
 Immediate consequences have the greatest effect.
 Consequences select any behavior that precedes them.
 Operant conditioning occurs automatically.
 6

Most stimulus changes that function as reinforcers or punishers can be described as either (a) a
new stimulus added to the environment or (b) an already present stimulus removed from the
environment.
Positive reinforcement: A response is followed by the presentation of a stimulus that results in
similar responses occurring more often.
Negative reinforcement: A response is followed by the withdrawal of a stimulus that results in
similar responses occurring more often.
The term aversive stimulus is often used to refer to stimulus conditions whose termination
functions as reinforcement.
Extinction (withholding all reinforcement for a previously reinforced behavior) produces a
decrease in response rate to the behavior’s pre-reinforcement level.
Positive Punishment: A response is followed by the presentation of a stimulus that results in
similar responses occurring less often.
Negative Punishment: A response is followed by the withdrawal of a stimulus that results in
similar responses occurring less often.
A principle of behavior describes a functional relation between behavior and one or more of its
controlling variables that has thorough generality across organisms, species, settings, and
behaviors.
A behavior change tactic is a technologically consistent method for changing behavior that has
been derived from one or more basic principles of behavior.
Unconditioned reinforcers and punishers function irrespective of any prior learning history.
Stimulus changes that function as conditioned reinforcers and punishers do so because of
previous pairing with other reinforcers or punishers.
One important function of motivating operations is altering the current value of stimulus changes
as reinforcement or punishment. For example, deprivation and satiation are motivating
operations that make food more or less effective as reinforcement.
A discriminated operant occurs more often under some antecedent conditions than it does under
others, an outcome called stimulus control.
Stimulus control refers to differential rates of operant responding observed in the presence or
absence of antecedent stimuli.
Antecedent stimuli acquire the ability to control operant behavior by having been paired with
certain consequences in the past.
 7

The Three-Term Contingency—antecedent, behavior, and consequence—is the basic unit of

analysis in the analysis of operant behavior.
If a reinforcer (or punisher) is contingent on a particular behavior, the behavior must be emitted
for the consequence to occur.
All applied behavior analysis procedures involve manipulation of one or more components of the
three-term contingency.
Recognizing the Complexity of Human Behavior
Humans are capable of acquiring a huge repertoire of behaviors. Response chains and verbal
behavior also make human behavior extremely complex.
The variables that govern human behavior are often highly complex. Many behaviors have
multiple causes.
Individual differences in histories of reinforcement and organic impairments also make the
analysis and control of human behavior difficult.
Applied behavior analysts are sometimes prevented from conducting an effective analysis of
behavior because of practical, logistical, financial, sociopolitical, legal, and/or ethical reasons.

Part 2 Selecting, Defining, and Measuring Behavior

Selecting and Defining Target Behaviors

Role of Assessment in Applied Behavior Analysis

Behavioral assessment involves using indirect, direct, and empirical methods to identify, define,
and determine the function of target behaviors.
Behavioral assessment consists of five phases or functions: (a) screening, (b) defining and
quantifying problems or goals, (c) pinpointing the target behavior(s) to be treated, (d) monitoring
progress, and (e) following up.
Before conducting a behavioral assessment, the behavior analyst must determine whether she has
the authority and permission, resources, and skills to assess and change the behavior. Past and
current records related to medical, educational, and historical events should be examined and
analyzed as part of a complete behavioral assessment.
Assessment Methods Used by Behavior Analysts
Subsumed under the three major methods for assessment information—indirect, direct, and
empirical—are (a) interviews, checklists/rating scales, (b) tests and direct observations, and (c)
 8

functional behavior analysis and reinforcer/punishment preference assessment methods,

respectively.
The client interview is used to determine the client’s description of problem behaviors or
achievement goals. What, when, and where questions are emphasized, focusing on the actual
behavior of the client and the responses of significant others to that behavior.
Questionnaires and needs assessment surveys are sometimes completed by the client to
supplement the information gathered in the interview.
Clients are sometimes asked to self-monitor certain situations or behaviors. Self-collected data
may be useful in selecting and defining target behaviors.
Significant others can also be interviewed to gather assessment information and, in some cases,
to find out whether they will be willing and able to assist in an intervention.
Direct observation with a behavior checklist that contains specific descriptions of various skills
can indicate possible target behaviors.
Anecdotal observation, also called ABC recording, yields a descriptive, temporally sequenced
account of all behaviors of interest and the antecedent conditions and consequences for those
behaviors as those events occur in the client’s natural environment.
Ecological assessment entails gathering a large amount of information about the person and the
environments in which that person lives and works (e.g., physiological conditions, physical
aspects of the environment, interactions with others, past reinforcement history). A complete
ecological assessment is neither necessary nor warranted for most applied behavior analysis
programs.
Reactivity refers to the effects that an assessment procedure has on the behavior being assessed.
Behavior analysts should use assessment methods that are as unobtrusive as possible, repeat
observations until apparent reactive effects subside, and take possible reactive effects into
account when interpreting the results of observations.
Choose assessment methods that produce reliable and valid results, conduct assessments
according to professional standards, and apply conservative analyses when interpreting results.
Assessing the Social Significance of Potential Target Behaviors
Target behaviors in applied behavior analysis must be socially significant behaviors that will
increase a person’s habilitation (adjustment, competence, and quality of life).
The relative social significance and habilitative value of a potential target behavior can be
clarified by viewing it in light of the following considerations:
 Will the behavior be reinforced in the person’s daily life? The relevance of behavior rule
requires that a target behavior produce reinforcement for the person in the
postintervention environment.
 Is the behavior a necessary prerequisite for a useful skill?
 9

 Will the behavior increase the person’s access to environments in which other important
behaviors can be learned or used?
 Will the behavior predispose others to interact with the person in a more appropriate and
supportive manner?
 Is the behavior a cusp or pivotal behavior? Behavioral cusps have sudden and dramatic
consequences that extend well beyond the idiosyncratic change itself because they expose
the person to new environments, reinforcers, contingencies, responses, and stimulus
controls. Learning a pivotal behavior produces corresponding modifications or
covariations in other untrained behaviors.
 Is the behavior age appropriate?
 Whenever a behavior is targeted for reduction or elimination, a desirable, adaptive
behavior must be selected to replace it.
 Does the behavior represent the actual problem or achievement goal, or is it only
indirectly related?
 A person’s verbal behavior should not be confused with the actual behavior of interest.
However, in some situations the client’s verbal behavior should be selected as the target
behavior because it is the behavior of interest.
 If a person’s goal is not a specific behavior, a target behavior(s) must be selected that will
produce the desired results or state.
Prioritizing Target Behaviors
Assessment often reveals more than one possible behavior or skill area for targeting.
Prioritization can be accomplished by rating potential target behavior against key questions
related to their relative danger, frequency, long-standing existence, potential for reinforcement,
relevance for future skill development and independent functioning, reduced negative attention
from others, likelihood of success, and cost.
Participation by the person whose behavior is to be changed, parents and/or other important
family members, staff, and administration in identifying and prioritizing target behaviors can
help reduce goal conflicts.
Defining Target Behaviors
Explicit, well-written target behavior definitions are necessary for researchers to accurately and
reliably measure the same response classes within and across studies or to aggregate, compare,
and interpret their data.
Good target behaviors definitions are necessary for practitioners to collect accurate and
believable data to guide ongoing program decisions, apply procedures consistently, and provide
accountability to clients, parents, and administrators.
Function-Based Definitions designate responses as members of the targeted response class
solely by their common effect on the environment.
 10

Topography-Based Definitions define instances of the targeted response class behavior by the
shape or form of the behavior.
A good definition must be objective, clear, and complete, and must discriminate between what is
and what is not an instance of the target behavior.
A target behavior definition is valid if it enables observers to capture every aspect of the behavior
that the “complainer” is concerned with and none other.
Setting Criteria for Behavior Change
A behavior change has social validity if it changes some aspect of the person’s life in an
important way.
Outcome criteria specifying the extent of behavior change desired or needed should be
determined before efforts to modify the target behavior begin.
Two approaches to determining socially validated performance criteria are (a) assessing the
performance of people judged to be highly competent and (b) experimentally manipulating
different levels of performance to determine which produces optimal results.

Measuring Behavior

Definition and Functions of Measurement in Applied Behavior Analysis

Measurement is the process of applying quantitative labels to observed properties of events using
a standard set of rules.
Measurement is how scientists operationalize empiricism.
Without measurement, all three levels of scientific knowledge—description, prediction, and
control—would be relegated to guesswork and subjective opinions.
Applied behavior analysts measure behavior to obtain answers to questions about the existence
and nature of functional relations between socially significant behavior and environmental
variables.
Practitioners measure behavior before and after treatment to evaluate the overall effects of
interventions (summative evaluation) and conduct frequent measures of behavior during
treatment (formative assessment) to guide decisions concerning the continuation, modification,
or termination of treatment.
Without frequent measures of the behavior targeted for intervention, practitioners may (a)
continue an ineffective treatment when no real behavior change occurred or (b) discontinue an
effective treatment because subjective judgment detects no improvement.
 11

Measurement also helps practitioners optimize their effectiveness; verify the legitimacy of
practices touted as “evidence based”; identify treatments based on pseudoscience, fad, fashion,
or ideology; be accountable to clients, consumers, employers, and society; and achieve ethical
standards.
Measurable Dimensions of Behavior
Because behavior occurs within and across time, it has three dimensional quantities: repeatability
(i.e., count), temporal extent (i.e., duration), and temporal locus (i.e., when behavior occurs).
These properties, alone and in combination, provide the basic and derived measures used by
applied behavior analysts.
 Count is a simple tally of the number of occurrences of a behavior.
 Rate is a ratio of count per observation period; it is expressed as count per standard unit
of time.
 Celeration is a measure of the change (acceleration or deceleration) in rate of responding
per unit of time.
 Duration is the amount of time from the onset to the end point of a behavior.
 Response Latency is a measure of the elapsed time between the onset of a stimulus and
the initiation of a subsequent response.
 Interresponse time (IRT) is the amount of time that elapses between two consecutive
instances of a response class.
 Percentage, a ratio formed by combining the same dimensional quantities, expresses the
proportional quantity of an event in terms of the number of times the event occurred per
100 opportunities that the event could have occurred.
 Trials-To-Criterion is a measure of the number of response opportunities needed to
achieve a predetermined level of performance.
Although form (i.e., topography) and intensity of responding (i.e., magnitude) are not
fundamental dimensional quantities of behavior, they are important quantitative parameters for
defining and verifying the occurrence of many response classes.
 Topography refers to the physical form or shape of a behavior.
 Magnitude refers to the force or intensity with which a response is emitted.
 12

Methods for Measuring Behavior

Event Recording encompasses a wide variety of procedures for detecting and recording the
number of times a behavior of interest is observed.
Time Sampling refers to a variety of methods for observing and recording behavior during
intervals or at specific moments in time.
Observers using Whole-Interval Recording divide the observation period into a series of equal
time intervals. At the end of each interval, they record whether the target behavior occurred
throughout the entire interval.
Observers using Partial-Interval Recording divide the observation period into a series of equal
time intervals. At the end of each interval, they record whether behavior occurred at any point
during the interval.
Observers using Momentary Time Sampling divide the observation period into a series of time
intervals. At the end of each interval, they record whether the target behavior is occurring at that
specific moment.
Planned Activity Check (PLACHECK) is a variation of momentary time sampling in which the
observer records whether each individual in a group is engaged in the target behavior.
Measuring Behavior by Permanent Products
Measuring behavior after it has occurred by measuring its effects on the environment is known as
measurement by permanent product.
Measurement of many behaviors can be accomplished with contrived permanent products.
Measurement by permanent product offers numerous advantages: The practitioner is free to do
other tasks; it enables the measurement of behaviors that occur at inconvenient or inaccessible
times and places; measurement may be more accurate, complete, and continuous; it facilitates the
collection of interobserver agreement and treatment integrity data; and it enables the
measurement of complex behaviors and multiple response classes.
If moment-to-moment treatment decisions must be made during the session, measurement by
permanent product may not be warranted.
Behaviors suitable for measurement with permanent products must meet two rules.
Rule 1: Each occurrence of the target behavior must produce the same permanent product.
Rule 2: The permanent product can be produced only by the target behavior.
 13

Measurement Tools
Low tech tools are functional, easy to use, inexpensive, and effective for research, education, and
treatment.
High-tech, digital and computer hardware and software systems for behavioral measurement and
data analysis have become increasingly sophisticated and easier to use.
Developers have produced data collection and analysis software for observational measurement
using laptops, handheld computers, personal digital assistants (PDAs), and desktop computers.
These measurement tools provide the researcher and practitioner with advantages of practical
value, efficiency, and ease of use.
Selecting a Measurement Method
When selecting a measurement method, consider the behavior change goals and expected
direction of behavior change, the relative ease of detecting occurrences of the behavior, the
environments where and times when the behavior will be measured, and the availability and
skills of personnel who will observe and record the behavior.
Improving and Assessing the Quality of Behavioral Measurement

Indicators of Trustworthy Measurement

To be most useful for science, measurement must be valid, accurate, and reliable.
Valid measurement in ABA encompasses three equally important elements:
1. Measuring directly a socially significant target behavior,
2. Measuring a dimension of the target behavior relevant to the question or concern about
the behavior
3. Ensuring that the data are representative of the behavior under conditions and during
times most relevant to the reason(s) for measuring it.
Measurement is accurate when observed values, the data produced by measuring an event, match
the true state, or true values, of the event.
Measurement is reliable when it yields the same values across repeated measurement of the same
event.
 14

Threats to Valid Measurement

Indirect measurement—measuring a behavior different from the behavior of interest—
threatens validity because it requires the researcher or practitioner to make inferences about the
relationship between the measures obtained and the actual behavior of interest.
A researcher who uses indirect measurement must provide evidence that the behavior measured
directly reflects, in some reliable and meaningful way, something about the behavior for which
the researcher wishes to draw conclusions.
Measuring a dimension of the behavior that is ill suited for, or irrelevant to, the reason for
measuring the behavior compromises validity.
Measurement artifacts are data that give an unwarranted or misleading picture of the behavior
because of the way measurement was conducted. Discontinuous measurement, poorly scheduled
observations, and insensitive or limiting measurement scales are common causes of measurement
artifacts.
Threats to Accurate and Reliable Measurement
Most investigations in applied behavior analysis use human observers to measure behavior, and
human error is the biggest threat to the accuracy and reliability of data.
Factors that contribute to measurement error include poorly designed measurement systems,
inadequate observer training, and expectations about what the data should look like.
Observers should receive systematic training and practice with the measurement system and
meet predetermined accuracy and reliability criteria before collecting data.
Observer Drift—unintended changes in the way an observer uses a measurement system over
the course of an investigation—can be minimized by booster training sessions and feedback on
the accuracy and reliability of measurement.
An observer’s expectations or knowledge about predicted or desired results can impair the
accuracy and reliability of data.
Observers should not receive feedback about the extent to which their data confirm or run
counter to hypothesized results or treatment goals.
Measurement Bias caused by observer expectations can be avoided by using naive observers.
Observer reactivity is measurement error caused by an observer’s awareness that others are
evaluating the data he reports.
 15

Assessing the Accuracy and Reliability of Behavioral Measurement

Researchers and practitioners who assess the accuracy of their data can (a) determine early in an
analysis whether the data are usable for making experimental or treatment decisions, (b) discover
and correct measurement errors, (c) detect consistent patterns of measurement error that can lead
to the overall improvement or calibration of the measurement system, and (d) communicate to
others the relative trustworthiness of the data.
Assessing the accuracy of measurement is a straightforward process of calculating the
correspondence of each measure, or datum, assessed to its true value.
True values for many behaviors of interest to applied behavior analysts are evident and
universally accepted or can be established conditionally by local context. True values for some
behaviors (e.g., cooperative play) are difficult because the process for determining a true value
must be different from the measurement procedures used to obtain the data one wishes to
compare to the true value.
Assessing the extent to which observers are reliably applying a valid and accurate measurement
system provides a useful indicator of the overall trustworthiness of the data.
Assessing the reliability of measurement requires a natural or contrived permanent product so the
observer can remeasure the same behavioral events.
Although high reliability does not confirm high accuracy, discovering a low level of reliability
signals that the data are suspect enough to be disregarded until problems in the measurement
system can be determined and repaired.
Using Interobserver Agreement to Assess Behavioral Measurement
The most commonly used indicator of measurement quality in ABA is interobserver agreement
(IOA), the degree to which two or more independent observers report the same observed values
after measuring the same events.
Researchers and practitioners use measures of IOA to (a) determine the competence of new
observers, (b) detect observer drift, (c) judge whether the definition of the target behavior is clear
and the system not too difficult to use, and (d) convince others of the relative believability of the
data.
Measuring IOA requires that two or more observers (a) use the same observation code and
measurement system, (b) observe and measure the same participant(s) and events, and (c)
observe and record the behavior independent of influence by other observers.
There are numerous techniques for calculating IOA, each of which provides a somewhat
different view of the extent and nature of agreement and disagreement between observers.
Percentage of agreement between observers is the most common convention for reporting IOA in
ABA.
 16

IOA for data obtained by event recording can be calculated by comparing (a) the total count
recorded by each observer per measurement period, (b) the counts tallied by each observer
during each of a series of smaller intervals of time within the measurement period, or (c) each
observer’s count of 1 or 0 on a trial-by-trial basis.
Total count IOA is the simplest and crudest indicator of IOA for event recording data, and exact
count-per-interval IOA is the most stringent for most data sets obtained by event recording.
IOA for data obtained by timing duration, response latency, or interresponse time (IRT) is
calculated in essentially the same ways as for event recording data.
Total duration IOA is computed by dividing the shorter of the two durations reported by the
observers by the longer duration. Mean duration-per-occurrence IOA is a more conservative and
usually more meaningful assessment of IOA for total duration data and should always be
calculated for duration-per-occurrence data.
Three techniques commonly used to calculate IOA for interval data are
1. Interval-By-Interval IOA
2. Scored-Interval IOA
3. Unscored-Interval IOA.
Because it is subject to random or accidental agreement between observers, interval-by-interval
IOA is likely to overestimate the degree of agreement between observers measuring behaviors
that occur at very low or very high rates.
Scored-interval IOA is recommended for behaviors that occur at relatively low frequencies;
unscored-interval IOA is recommended for behaviors that occur at relatively high frequencies.
IOA assessments should occur during each condition and phase of a study and be distributed
across days of the week, times of day, settings, and observers.
Researchers should obtain and report IOA at the same levels at which they report and discuss the
results of their study.
More stringent and conservative IOA methods should be used over methods that may
overestimate agreement as a result of chance.
The convention for acceptable IOA has been a minimum of 80%, but there can be no set
criterion. The nature of the behavior being measured, and the degree of behavior change revealed
by the data must be considered when determining an acceptable level of IOA.
IOA scores can be reported in narrative, table, and graphic form.
Researchers can use multiple indices to assess the quality of their data (e.g., accuracy plus IOA,
reliability plus IOA).
 17

Reference

Cooper, J. O., Heron, T. E., Heward, W. L. (20190626). Applied Behavior Analysis, 3rd Edition.

[[VitalSource Bookshelf version]]. Retrieved from vbk://9780134798783

18