AB only design
A faulty within-subject experimental design composed of a baseline condition and an experimental condition only. Behavioral changes based on time coincidences cannot be ruled out with this method. No valid conclusions can be drawn from this method.
An appropriate single-subject design in which assessments are made across Baseline, Treatment, and a Return to Baseline.
An extension of the ABA design that requires a final return to Treatment. Considered more ethical if treatment produces more acceptable or more adaptive behavior.
Baseline, Treatment1 and Treatment 2. Sometimes referred to as an additive design. Lacks replication and hence, is considered inappropriate.
The extent to which observed values approximate the true values (aka the events that actually occurred).
Repetition of behavioral phenomena under different conditions across different ﬁelds of science.
alternating treatments design
An appropriate single-subject design in which several treatments are assessed repeatedly and in random order. Rapid and random alternations of treatments make it difficult for the participants to know which condition they are in.
A faulty within-subject experimental design composed of an experimental condition only. Appropriate baseline data are not gathered. No valid conclusions can be drawn from this method.
Treatment, Baseline and a Return to treatment. This intervention begins with treatment due to the seriousness of the behavior under study.
Any condition or phase of an experiment in which the independent variable (e.g. intervention or treatment) is not present.
Very precise operational definitions that are objective, that have face validity and that are reliable. Typically, the behavior must be distinguished from other events in a qualitative and quantitative fashion.
Variations in features of responding within a single response class. These responses are functionally equivalent but may vary in efficacy and/or form.
A method of arranging comparisons between control and experimental conditions in which different groups of subjects are exposed to control and experimental conditions so that the data represent the combined performance of individual participants who have experienced only one of the conditions.
A reference to being unaware of the goals of an experiment, the nature of experimental or control conditions, or the outcomes of each condition. May apply to the experimenter, observers, or participants. If both experimenter and participants are blind, it is called a double-blind experiment.
A brief period of responding deﬁned by the relatively frequent occurrence of one or more speciﬁc response classes and which is distinguished from other relatively extended periods in which the target responses do not occur (e.g. self stimulation).
Evaluating the accuracy and reliability of data produced by a measurement procedure and, if necessary, using these ﬁndings to improve the procedure so that it meets desired standards.
changing criterion design
A single subject design using gradually increasing quantitative approximations (steps) to the final target performance. Each experimental phase is additive. Reversals may be employed.
A data collection technique in which all instances of the target behavior are observed and recorded.
A graph showing the increasing number of responses over time. Each response corresponds to a single "tic" on the graph. The faster the responding, the steeper the slope.
Dead Man's Test
An informal test of whether a particular event is a behavior. The test is that if a dead man can do it, it is not behavior.
diffusion of treatment
An experimental confound in which the intended differences between baseline and treatment are inadvertently blended. (e.g. low levels of reinforcement are continued during extinction).
Measurement practices in which the events measured are the same as those about which conclusions will be drawn.
Methods in which all target responses are not necessarily detected and recorded.
A method of data collection in which frequency counts are gathered on one or more behaviors (occurred/not occurred; correct/incorrect; appropriate/inappropriate).
A series of independent opportunities for the subject to emit a single response to a specific discriminative stimulus. Successive trials are separated by an inter-trial interval.
The elapsed time between the beginning and ending of an event, such as a single response (e.g. crying) or a response bout (e.g. SIB).
A procedure for calculating interobserver agreement that involves dividing the observation period into intervals in which two observers record the actual number of responses. In order to obtain percent agreement, only intervals in which the two observers agreed on the exact count are considered agreements.
The management or control of different variables in a study, including the independent variable and extraneous variables.
Arrangement of control and treatment conditions that permit comparisons that help to identify the effects of the independent variable on the dependent variable.
Environmental events that are not of interest to the researcher but that may influence the participant's behavior is ways that obscure the effects of the independent variable.
The assumption that behavioral variability can be explained in term of variation(s) in biological or environmental phenomena.
focused line projection
A visual analysis method for drawing the best fit line that relies on locating equal data points above and below a line. Slightly more reliable than free-hand line projection.
An experimentally determined relation that shows that the dependent variable depends on or is a function of the independent variable and nothing else.
The meaningfulness of experimental interpretations under circumstances different from those that generated the data.
The mathematical result of combining the data from multiple individuals who are related in some way (e.g., exposure to an experimental condition or applied intervention). Does not refer to a natural phenomenon distinct from the mere sum of behaviors from various individuals.
The degree to which relevant individuals (e.g. employers, caregivers, parents) believe that targeted changes in behavior will produce meaningful changes in the client's proficiency at a given functional skill set.
A plotting convention referring to days on which the target behavior could have occurred but was not measured. Data points bracketing such days should be connected in line graphs.
A schedule of observation that samples from the population of responses in the deﬁned class.
Measurement practices in which the events measured are not the same as those about which conclusions will be drawn.
A procedure for enhancing the believability of data that involves comparing simultaneous but independent observations from two or more observers. Provides no information about accuracy or reliability.
interresponse time (IRT)
The time elapsing between two successive responses. Measured as time per cycle.
A procedure for calculating interobserver agreement when interval recording or time sampling is used. Each interval scored by two observers is counted as an agreement, and each interval that is scored by neither observer is also called an agreement. Intervals for which only one observer scored the behavior are counted as disagreements.
Observation periods are divided into equal intervals which are scored according to some rule if the target behavior occurs.
The assumption that variability in behavior is in one way or another inherent or built into the nature of organisms.
Measurement of the elapsed time between the onset of a stimulus and the occurrence of behavior.
linear interval scale
A measurement scale that allows descriptions of events that show how much they differ in terms of equal intervals between values.
local response rate
Rate of responding across a period of time that is smaller than that for which the overall responses rate is determined.
logarithmic interval scale
A measurement scale that allows descriptions of events that show how much they differ in terms of equal ratios between values.
A horizontal graphical line that represents the quantitative mean for a given condition.
A horizontal line that represents the most typical performance within a given condition. Best used when the mean is not representative of actual performance.
momentary time sampling
A discontinuous observation procedure in which observation periods are divided into equal intervals, but the observer only notes the momentary status of the target behavior at the end of each interval.
A variation of a reversal design that exposes a participant ﬁrst to one condition and then to another in some form of repeated alternation. Also called an alternating-treatment design.
multiple baseline design
A within-subject design that uses two or more baselines in a coordinated way to allow control-treatment comparisons both within and across baselines.
multiple probe design
A variation of the multiple baseline design is which regular, consistent measures are taken before and after treatment until the behavior becomes stable. Occasional assessments of other behaviors, locations or subjects, provide indicators of ongoing stable behavior. Useful when continuous/daily data collection is not possible.
A method for determining the role of contingency or the role of motivation in behavior change. If RFT is delivered in a noncontingent manner and the behavior fails to improve, it is not the increased reinforcement density alone that is influencing behavior, but rather the reinforcement contingency.
no chance day
A plotting convention referring to days on which the target behavior could not have occurred. Data points bracketing such days should not be connected in line graphs.
Values resulting from observation and recording procedures used to collect the data for a study
A gradual shift in the accuracy of an observer's performance, often for unknown reasons.
A conservative approach to calculating interobserver agreement when interval recording or time sampling is used that involves calculating and reporting agreement separately for both occurrences (scored intervals) and nonoccurrences (unscored intervals).
Using the physical product of behavior as the measureable feature of the behavior occurring. Only useful if there is a reliable one-to-one correspondence between behavior and its product.
overall response rate
The average rate of responding in a given condition. Usually represented in a histogram or on a line graph as the mean-level line.
partial interval recording
An interval is scored if at least one target response (or even part of a response) occurs during the interval.
A measurement system that expresses the amount of a behavior as a measure relative to the total opportunities to respond (e.g. correct or erroneous responses).
Investigative methods that do not eliminate the influence of extraneous factors. Referred to as faulty designs if used as the only experimental method. Includes case studies.
The ability to anticipate the outcome of a future measurement that comes as a result of repeated, systematic observations of a behavior or event over time.
Temporary changes in behavior (for observers or clients) that are the direct result of knowing that they are being observed or assessed.
A property of events that can recur. In the study of behavior, a property of a class of functionally equivalent responses. Described in terms of countability.
A Pavlovian response (e.g. reflex) elicited by particular unconditioned or conditioned antecedent stimuli.
Tangible environmental effects of responding that are more than transitory in duration.
A within-subject experimental design involving a pair of control and experimental conditions in which one or both conditions repeat at least once.
Used to indicate discontinuity in the progression of the scale of measure on the abscissa or the ordinate.
An effect of a participant's behavior resulting from exposure to a prior condition.
single-subject interaction designs
Any appropriate single-subject design in which some assessments conditions are included that combine treatments. These designs must have sufficient conditions to establish stimulus control and to rule out alternative explanations. E.g. A B A B BC B BC
The use of normative data to determine treatment efficacy. Selection of a normative group may require sensitivity to age, race, or disability.
A variety of methods to assess whether an intervention has produced behavioral changes of clinical or applied importance.
A nonparametric technique used to calculate the best fit line that relies on arithmetic calculations for each half of the data set.
steady state behavior
A pattern of responding that shows relatively little variation in its measured dimensional quantities over some period of time.
The experimental demonstration of control over behavior by the different levels of the IV. Typically requires multiple replications and reversals.
An additional assessment of the efficacy of a behavioral intervention by people who are in a special position to judge (e.g. teachers, parents, caregivers). If a client has made meaningful changes in behavior, it should be apparent to those who work with the client.
A statistical analysis used in single subject designs that provide information about levels and trends across experimental phases.
A procedure for calculating interobserver agreement typically used with dimensional quantities such as count, duration, and latency; involves summing the total count for each of two observers, dividing the smaller total by the larger total, and multiplying the result by 100 to arrive at the percent agreement.
The pattern of responding involving the change from one steady state to a different steady state resulting from a change in the experimental condition (e.g. transition from baseline to treatment). Typically not included in the final description of behavior under the given conditions.
A pattern of responding involving a deviation from a steady state that ends in a return to the same steady state within the same condition. May be cyclical in nature.
A condition or phase in which the independent variable (e.g. intervention) is present.
The extent to which the independent variable is consistently implemented as it was designed.
The direction of a data path that may include increasing, decreasing, stable, variable or flat patterns.
A graphed line that represents the direction of a data path within a given condition (aka line of progress).
A measurement scale that establishes the number of trials required to meet a pre-established response requirement. Comparisons are made across responses in the same or similar response classes (e.g. response generalization, stimulus discriminations).
Values resulting from special observational and recording procedures that are somewhat different from those used to collect the data being evaluated and that involve special efforts to minimize error. (e.g. derived from video or audio recordings).
unit of analysis
A constituent part of a whole phenomenon that serves as a basis for experimental study. In the study of behavior, the unit of analysis is the response class.
The extent to which observed values represent the events they are supposed to represent and that will be the focus of interpretation.
A systematic and comprehensive (non-statistical) approach to the interpretation of graphed data used to determine treatment efficacy.
whole interval recording
An interval is scored if the target behavior occurs without ceasing throughout the entire interval.
Repetition of an entire phase or condition during the course of an experiment.