Experimental behavior analysis

The Experimental behavior analysis (Engl. Experimental Analysis of Behavior , EAB) is one of BF Skinner reasoned scientific research tradition, originally within the psychology , meanwhile, the research also includes questions of biology , medicine and economics . Based on the scientific theory of radical behaviorism , experimental behavior analysis examines the functional relationships between the behavior of animals and humans and the environment . It forms the basic scientific aspect of behavior analysis .

Basics

history

Experimental behavior analysis was established through the pioneering scientific work of BF Skinner . The publication of Skinner's work " The Behavior of Organisms: An Experimental Analysis " in 1938, in which he described the basics of operant learning and its effects on organisms , is considered a milestone in the development of experimental behaviors . Based on Skinner's work, laboratories have been established around the world to study the behavior of organisms. One of the best known was the "Harvard Pigeon Lab" which Skinner founded in 1948 at Harvard University . Also in 1948 the first conference on experimental behavior analysis took place in Bloomington . In 1949, Paul Fuller's article “Operant Conditioning of a Vegetative Human Organism” appeared, probably the first article to deal with operant conditioning on human subjects.

1958 was scientific journal the Journal of the Experimental Analysis of Behavior founded, which is considered the most important organ of publication in the field of experimental behavior analysis. In 1968 the Journal of Applied Behavior Analysis was founded , which deals with the implementation of the findings from experimental behavior analysis in practice as an applied science .

Today most of the behavioral analysts worldwide are united in the Association for Behavior Analysis International , of which there is also a European section ( European Association for Behavior Analysis ).

The techniques and findings developed in the context of experimental behavior analysis (e.g. Skinner box ) are also used in other scientific areas such as biology or medicine .

aims

Experimental behavior analysis deals with the functional relationships between the behavior of humans and animals and the environment. The aim is to describe, explain, predict and control behavior. One term to describe this goal is “functional contextualism ”. Behavior can only be understood through the function of the behavior and in which context the behavior occurs.

Inductive approach

Methodologically , the experimental behavior analysis follows a data-based inductive approach. That is, behavioral analysts try to draw theoretical principles from data obtained through experiment. This procedure does not mean that the behavioral analyst has no assumptions about the outcome of the experiment before the experiment. However, the experimental behavior analysis dispenses with the deductive- mathematical derivation of formal theories, such as those advocated by Clark Hull .

Differences to other experimental psychological approaches

→ Main article : Psychological experiment

Experimental behavior analysis takes on a certain special role among the psychological research traditions. These differences from many other psychological approaches are expressed both in the subject matter and in the methodology .

Most psychological experiments deal with hypothetical mental constructs (e.g. will , attitudes , etc.). These hypothetical constructs do not play a decisive role in experimental behavior analysis. Behavior should be explained without resorting to hypothetical constructs. This is sometimes misunderstood in the sense that behavior analysis denies internal psychological processes. This is incorrect. Internal events such as thinking or problem solving are also examples of behavior for behavior analysis. These are called private events, as they can only be perceived by one person in principle. However, it is criticized that, although the validity of private events is affirmed in the behavioral analysis, on the other hand these are largely ignored in practice.

Important principles

In its history, experimental behavior analysis has discovered some fundamental principles without which it is difficult to understand the research carried out in the context of experimental behavior analysis.

Function and topography of behavior

An important distinction in experimental behavior analysis concerns the topography of behavior and its function.

The function of behavior describes the connection between behavior and an environmental event. For example, when a rat presses a lever in a Skinner box, it represents a functional connection. The rat presses the lever (the specific behavior) and receives a food pill (the environmental event). One can now observe that the rat will operate the lever more often if it receives a food pill every time, it is reinforced for it . There is a functional relationship between pressing the lever and ejecting a food pill. In behavior analysis, behaviors that have the same function are grouped into functional classes.

The topography of behavior describes the behavior that can be observed from outside. For example, whether the rat operates the lever with one paw or with both paws.

For experimental behavior analysis, the functions of behavior and the functional classes derived from them are usually of greater interest than the topography of behavior. With the help of the functional relationships, current and future behavior can be explained, predicted and controlled.

Respondent and operant conditioning

Respondent conditioning

→ Main article : Classical conditioning

Respondent conditioning, also known as classical conditioning, describes learning that is based on a relatively simple stimulus-response mechanism. Although the principle of respondent conditioning has been known for over 100 years, there is still a lot of research in this field.

Operant conditioning

→ Main article : Instrumental and operant conditioning

The "heart" of experimental behavior analysis is operant conditioning , learning through consequences. All essential principles of operant conditioning such as reinforcement , punishment or reinforcement plans were discovered in the context of experimental behavior analysis.

Discriminative cues and contingency

A discriminative cue is a stimulus that signals a historical connection between a behavior and an reinforcing or punishing consequence. A classic example from experimental behavior analysis: pigeons are trained to peck for a disc to receive food only when a green light is on, but not when a red light is on. The light serves here as a discriminative reference stimulus.

Contingency is a direct connection between a behavior and a consequence. A certain consequence usually has to follow a behavior contingently in order to be able to exercise a function. This means that a direct connection between the behavior and the consequences is necessary. So that discriminative cues and consequences can permanently exert their function on the behavior, they must occur in contiguity with the behavior.

Rule-guided and contingency-shaped behavior

There is an important difference between rule-based and contingency-shaped behavior . While behavior in contiguous behavior is directly formed by the consequences following the behavior, people in particular can learn behavior through rule-based behavior - thanks to verbal rules - without having to have carried it out themselves. This ability significantly expands human behavioral repertoire and, according to behavioral analysts, is the source of many human cultural achievements.

Functional analysis

It is now possible to set up a functional analysis based on the principles mentioned above. A functional analysis is a model for describing operant processes. Functional analysis is also known under the name ABC model. The three letters stand for the three different parts of a functional analysis:

A → B → C

A (antecedent) : A denotes the preceding (antecedent) functions of the behavior. This includes all stimuli and events that precede and influence behavior, e.g. B. discriminative stimuli or reinforcing consequences from past actions.

B (behavior) : B denotes behavior and thus everything an organism does. Both public (visible to outsiders) and private events such as thinking fall under it.

C (consequence): C denotes all consequences of the behavior. These can have reinforcing or punishing functions, ie they can increase or decrease the likelihood of the behavior occurring in the future.

methodology

Single-subject research (case-by-case analysis)

Experimental behavior analysis deals with the individual behavior of organisms. For this reason, the statistical group comparisons with inferential statistical methods such as t-tests used by other psychological research traditions are rarely used. For the experimental behavior analysis, group comparisons with a large number of participants are unfavorable for psychological questions, since in group comparisons the uniqueness of the individuals takes a back seat in favor of statistical average values. Since most psychological theories deal with individuals, it is also unclear to what extent an individual individual can be inferred from the level of an examined group. Despite skepticism about group comparisons, these methods are also used by behavioral analysts, especially in applied science areas.

The modus operandi for experimental behavior analysis represents single-subject research (often also called single-subject design or individual case analysis). This refers to various methods of obtaining data from the individual behavior of one or more test persons. In single-subject research, measurements of the behavior of a subject are repeatedly carried out before, during and after an experimental manipulation . If the behavior of the test person changes after the introduction of an experimental manipulation, it is assumed that the experimental manipulation has an effect on the behavior. Various experimental designs are used to demonstrate this effect.

In the experimental behavior analysis, special emphasis is placed on the replicability of the experiments. This is done, for example, by each participant serving as his own control group. Overall, the experimental approach in behavior analysis is based more on the example of biology and less on that of psychology .

Basic principle

The overriding goal of experimental behavior analysis, as in general scientific studies, is to describe and explain relationships between variables. Various experimental designs are available for this purpose. Experimental designs are characterized by two criteria:

1) The variation of at least one independent variable (UV) and the investigation of the effect of the UV on a dependent variable (AV).

2) The exclusion of alternative explanations for changes in the AV. This includes in particular the control of interfering variables . The Skinner-Box, for example, enables the control of interfering variables, since the environment can be controlled very precisely in this.

Variation of UV in behavior analysis

Experimental group studies usually compare an intervention group and a control group at two measurement times with regard to the interesting target behavior. Behavioral analysts, on the other hand, examine the effect of an experimental manipulation on a few or even only on a single subject. The variation of the UV arises from the introduction of an experimental manipulation. Instead of two measurements on many people, many measurements are carried out on one person in behavior analysis. Repeated measurements are the central element of single-subject research. They are necessary above all because the features to be recorded are often subject to fluctuations or the measurements themselves can be incorrect. In the case of statistical group comparisons, on the other hand, only two measurement times are sufficient, since measurement errors and fluctuations are averaged over people.

Internal validity threats

There are several threats to internal validity for a case-by-case analysis .

Internal validity threats

Disturbance variable	description	control
history	Events that represent in the course of the investigation and exert an unwanted influence on the test subjects.	Use of different designs Animals as test subjects
maturation	Subjects change their behavior solely because of Maturation processes.	Use of different designs Animals as test subjects
Instrumentation	arises especially when observers change their assessment criteria during the Change examination and, for example, interpret behavior differently.	Multiple observers automatic recording of behavior clear definition of the behavior to be examined
Diffusion of treatment	describes the insufficient separation between the intervention and non-intervention phases z. B. if interventions in an ABA plan are continued during the non-intervention phase.	Clear separation between the phases

Data collection

An appropriate definition of the AV is important for the quality of the data collection. Behavioral definitions are used for this in behavior analysis. A behavioral definition indicates which behavior is to be measured. For reasons of objectivity , this usually means observable behavior. The recording of behavioral consequences (reinforcement or punishment) or physiological variables (e.g. heartbeat ) are also part of the tools of the behavioral analyst. Various aspects such as duration or frequency can be of interest from all these measurable variables.

In addition to the precise definition of the AV, it is also important that it can be reliably recorded and that it is change-sensitive. Change-sensitive means that the selected variables can adequately depict changes due to the intervention. For this purpose, continuous scales are usually used instead of dichotomous scales. For example, to answer the question of whether a punishment plan leads to a reduction in the behavior rate, it makes more sense to continuously record the behavior instead of just recording whether behavior occurs or not (dichotomous recording). If, for example, the introduction of a punishment plan were to reduce behavior from eleven times an hour to three times an hour, the success of the intervention would not be mapped using a dichotomous recording.

Technical devices for data collection play an important role in experimental behavior analysis. It would be very laborious and labor-intensive to observe the behavior of an experimental animal for days and to record the behavior shown manually. In addition, a mechanical data collection increases the reliability of the measurements. All modern Skinner boxes have an automatic recording of behavior such as pecking at a disc (if pigeons are the test animals). In addition, video recordings are often used for acquisition.

Designs

In experimental behavior analysis, various experimental designs have been developed. These can be used for research with humans as well as with animals. The core of all these designs is the creation of a chain of measurements by a test person for at least two different experimental conditions. First, the basic rate (eng. Baseline) of behavior is recorded in a measurement chain. As a rule, it is a question of behavior frequencies per unit of time, which are usually displayed graphically as a frequency polygon. The establishment of a baseline has three important goals: description, prognosis and comparison. Description means that the behavior is described in its "basic form", without experimental manipulation. From this, a prognosis can be derived about what effect an experimental manipulation would have on target behavior. The comparison of this prognosis with the data actually collected in the end forms the basis for assessing the intervention effect.

AB design

An example of an AB design with fictitious values. The measurement times are plotted on the x-axis and the measured values ​​on the y-axis.

In an AB design, in a first phase A, the behavior under investigation is repeatedly measured in several sessions without experimental manipulation, and the basic rate A is created from this data. The behavior is usually measured until the data no longer shows a trend and smooth curves form.

Then the experimental manipulation (intervention) is carried out in phase B and the behavior is measured again repeatedly in several sessions. Intervention phase B is created from this data.

In principle, the researcher can now use the frequency polygon to determine whether the behavior has changed after the experimental manipulation. From the difference between the measurements from phase A and B, an experimental effect of the manipulation can now be concluded.

The advantage of an AB design is its easy implementation and ethical harmlessness. The disadvantage is its limited internal validity. Even if there are clear trends within the data, it cannot be ruled out that the changes are due to confounding variables. To increase the validity, further designs were developed.

ABA design

An example of an ABA design with fictitious values. The measurement times are plotted on the x-axis and the measured values ​​on the y-axis.

An ABA design is an extension of the AB design and offers increased security as to whether the experimental manipulation is really responsible for the change in behavior.

As in the AB design, the basic rate A and the intervention rate B are measured here. The experimental manipulation is then reversed (e.g. the intervention is overridden) and a new base rate A is recorded. If the base rate now falls back to or close to the original level before the introduction of the manipulation, it is considered quite likely that the experimental manipulation is responsible for the behavior change.

The advantage of an ABA design is the very high internal validity, since interfering variables (especially history and maturation) can be well controlled. The disadvantage of an ABA design is primarily ethical. By removing a successful intervention, depending on the behavior examined, one may accept a deterioration in the target behavior. This is particularly problematic in experiments with human subjects.

ABAB design

An example of an ABAB design with fictitious values. The measurement times are plotted on the x-axis and the measured values ​​on the y-axis.

An ABAB design provides even better information about whether an experimental manipulation was successful. Here all phases of the ABA design are run through and then the experimental manipulation is introduced a second time. If this new manipulation leads again to the same or sufficiently similar intervention rate as in the first manipulation, this is an even stronger indication that the experimental manipulation is responsible for the change in behavior.

In principle, the designs can be (theoretically) extended to infinity. So one can collect many base and intervention rates from AB * N.

Multiple base rates (multiple baseline)

This design can be used when withdrawing an intervention is not possible or ethically unjustifiable. For example, when measuring a psychotherapeutic intervention. In addition, with this design the change in several behaviors due to manipulation can be measured simultaneously.

With a multiple base rate, an intervention for different behaviors is introduced with a time delay. In this way you can see whether only the behavior changes for which the intervention was introduced.

The advantage of a multiple baseline design is its high internal validity. In addition, no reverse phase is necessary, which takes into account the ethical concerns about ABA designs. The disadvantage of this design form is that it is very time-consuming and involves a large number of measurements.

Data evaluation

Visual inspection

The graphical representation of the measured values ​​in the form of frequency polygons enables a visual inspection of the intervention effects. The times of observation are plotted on the abscissa in the diagram and the AV values ​​on the ordinate. In addition, the baseline and intervention phases are marked.

By analyzing the differences between the baseline and the intervention phase, possible effects can be identified. The advantage of this method is that large effects are very noticeable and directly visible. In cases in which possible effects do not "jump straight into the eye", various criteria can be used:

criteria	description
Changes in mean	Here the mean value of the individual phases is used as a criterion. Are there clear differences in mean values between the individual phases, this is a good indicator of an intervention effect
Changes in level	Changes in the level are clear jumps when introducing and removing an intervention. If they occur during phase changes, they are additional evidence of the effectiveness of the intervention
Changes in trend	The "direction" of the behavior in the graph d is called the trend. H. takes the behavior to or from. In order to make the trend visible, trend lines are usually drawn for the baseline and intervention phase.
latency	Latency is the time in which there are no changes in behavior after the introduction or removal of an intervention respectively. A short latency speaks for the effectiveness of the intervention. On the other hand, a long latency indicates that others Variables are responsible for behavior change.
Moving average values	With the help of the calculation of a moving average, a clearer structure can be brought into the data. This makes it easier to identify trends. The prerequisite is the availability of a sufficient number of measured values.

Using the criteria mentioned, it is possible to draw a complex judgment about the effectiveness of the intervention. It is also possible to determine peculiarities and abnormalities in the course of the examination which might have been lost in a standardized statistical evaluation.

Disadvantages of a visual inspection are limitations in objectivity and reliability. So it is quite possible that different evaluators come to different assessments with the same data material based on visual inspection.

statistical evaluation

The visual inspection can be supplemented by statistical methods.

Percentage of non-overlapping data points (PND)

The PND is a descriptive statistical measure for the quantification of an intervention effect. The logic behind this calculation is that the less the values ​​of the baseline and intervention phases overlap, the greater your intervention effect. The PND represents a descriptive measure and is therefore not tied to any interference statistical prerequisites such as normal distribution.

${\displaystyle PND=\quad {\frac {{\text{AnzahlBPhase}}<{\text{min.AnzahlAPhase}}}{\text{AnzahlBphase}}}\cdot 100\,\%}$

The higher the percentage, the stronger the intervention effect.

The great advantage of the PND is its simple calculation and interpretation. However, the PND does not provide any meaningful results if the baseline shows strong fluctuations.

Selected research fields in experimental behavior analysis

Stimulus control

Research on stimulus control deals with the question of how behavior is influenced by previous stimuli. How do living things differentiate between different stimuli, how do they react to them, and in what way is their behavior influenced by them? Behavioral analysts have demonstrated remarkable abilities in this area, for example in pigeons , which were previously viewed as genuinely human abilities. For example, pigeons learned to differentiate between pictures by the painters Claude Monet and Pablo Picasso and even generalized this to pictures by other artists. In another experiment, researchers were able to train pigeons to pass the mirror test.

Voting Behavior and Matching Law

Behavioral analysts study the choices made by humans and animals, especially when they can choose between two or more behaviors. One of the most important discovered as part of the experimental behavior analysis principles is the Matching Law of Richard Herrnstein . The matching law is a mathematical theory. The basic statement is that if two behavior alternatives exist at the same time, the relative proportion of a behavior is equal to the relative proportion of reinforcement. There is extensive research on matching law, and it has been proven for both humans and animals.

Behavioral Pharmacology

Behavioral Pharmacology is interdisciplinary research involving both behavioral analysts and pharmacologists . In contrast to psychopharmacology , behavioral pharmacology is not about the effect of psychoactive substances on the brain, but about their effects on the behavior of humans and animals. For example, the effect of drugs on operant conditioning is being researched.

Rule-based behavior

An important principle that distinguishes human from animal behavior is the ability of humans to obey rules. In this way, people can acquire a repertoire of behavior without performing the behavior themselves. Behavioral analysts study the influence of rules on human behavior. For example, the question of whether rule-based or contingency-based behavior has the greater influence in certain situations.

literature

Introductory textbooks

• Christoph Bördlein: Introduction to behavior analysis. Alibri, Aschaffenburg 2015, ISBN 978-3-86569-232-0 .
• Gregory J. Madden (Ed.): APA handbook of behavior analysis. Vol. 1: Methods and principles. American Psychological Association, Washington DC 2013, ISBN 978-1-4338-1112-8 .
• James E. Mazur: Learning and Behavior. Pearson Studium, Hallbergmoos 2006, ISBN 3-8273-7218-6 .

Web links

• Cambridge Center for Behavioral Studies
• Association for Behavior Analysis International
• European Association for Behavior Analysis
• Association for Behavior Analysis Germany - German Society for Behavioral Analysis
• Journal of the Experimental Analysis of Behavior and Journal of Applied Behavior Analysis

Individual evidence

1. ^ ^{A b} Susan M. Schneider: The Science of Consequences . Prometheus, New York 2012, ISBN 978-1-61614-662-7 , pp. 13 .
2. ↑ RJ Gerrig, PG Zimbardo: Psychology . 18th edition. Pearson, Munich 2008, ISBN 978-3-8273-7275-8 , pp. 123 .
3. ↑ Christoph Bördlein: Introduction to behavior analysis . 1st edition. Alibri, Aschaffenburg 2015, ISBN 978-3-86569-232-0 , pp. 12 .
4. ↑ Wolfgang Schönpflug: History and systematics of psychology . 3. Edition. Beltz, Weinheim 2013, ISBN 978-3-621-28029-7 , pp. 302 .
5. ^ BF Skinner: The behavior of organisms: an experimental analysis . Appleton-Century, Oxford 1938.
6. ↑ K Lattal: A Tribute To The Harvard Pigeon Lab, 1948-1998. In: Journal of the Experimental Analysis of Behavior . tape 77 , no. 3 , 2002, p. 301 , doi : 10.1901 / jeab.2002.77-301 , PMC 1284882 (free full text).
7. ↑ James A. Dinsmoor: A visit to Bloomington: The first Conference on the Experimental Analysis of Behavior . In: Journal of the Experimental Analysis of Behavior . tape 48 , no. 3 , November 1, 1987, pp. 441–445 , doi : 10.1901 / jeab.1987.48-441 , PMID 16812505 , PMC 1338767 (free full text).
8. ^ Paul R. Fuller: Operant Conditioning of a Vegetative Human Organism . In: The American Journal of Psychology . tape 62 , no. 4 , 1949, pp. 587-590 , doi : 10.2307 / 1418565 .
9. ^ Edward K. Morris, Deborah E. Altus, Nathaniel G. Smith: A Study in the Founding of Applied Behavior Analysis Through Its Publications . In: The Behavior Analyst . tape 36 , no. 1 , January 1, 2013, p. 73-107 , PMID 25729133 , PMC 3640891 (free full text).
10. ^ BF Skinner: Antecedents . In: Journal of the Experimental Analysis of Behavior . tape 48 , no. 3 , November 1, 1987, pp. 447-448 , doi : 10.1901 / jeab.1987.48-447 , PMID 16812506 , PMC 1338768 (free full text).
11. ↑ ^{a b} B. F. Skinner: What is the experimental analysis of behavior? In: Journal of the Experimental Analysis of Behavior . tape 9 , no. 3 , May 1, 1966, pp. 213-218 , doi : 10.1901 / jeab.1966.9-213 , PMID 16811287 , PMC 1338181 (free full text).
12. ^ E. V Gifford, SC Hayes: Functional contextualism: A pragmatic philosophy for behavioral science . In: William O'Donohue (Ed.): Handbook of Behaviorism . 1st edition. Academic Press, London 1999, pp. 285 .
13. ^ Mecca Chiesa: Radical behaviorism: The philosophy and the science . 1st edition. Authors Cooperative, Boston 1994, ISBN 0-9623311-4-7 , p. 45 .
14. ^ Murray Sidman: Tactics of Scientific Research: Evaluating Experimental Data in Psychology . Authors Cooperative, Boston 1960, ISBN 0-9623311-0-4 , p. 50 .
15. ^ A. Charles Catania: Learning . 5th edition. Sloan Publishing, Cornwall-on-Hudson 2013, ISBN 978-1-59738-023-2 , pp. 20 .
16. ^ Nicola Döring, Jürgen Bortz, Sandra Pöschl: Research methods and evaluation in the social and human sciences . 5th edition. Springer, Heidelberg 2015, ISBN 978-3-642-41088-8 , pp. 223 .
17. ^ BF Skinner: About Behaviorism . 1st edition. Vintage, New York 1976, ISBN 0-394-71618-3 , pp. 228 .
18. ↑ K. Geoffrey White, Dianne McCarthy, Edmund Fantino: Cognition and behavior analysis . In: Journal of the Experimental Analysis of Behavior . tape 52 , no. 3 , November 1, 1989, pp. 197–198 , doi : 10.1901 / jeab.1989.52-197 , PMID 16812594 , PMC 1339174 (free full text).
19. ^ P. Friman: Behavior Analysis Of Private Events Is Possible, Progressive, And Nondualistic: A Response To Lamal. In: Journal of Applied Behavior Analysis . tape 31 , no. 4 , 1998, pp. 707-708 , doi : 10.1901 / jaba.1998.31-707 , PMC 1284169 (free full text).
20. ^ BF Skinner: An operant analysis of problem solving . In: Behavioral and Brain Sciences . tape 7 , no. 4 , December 1, 1984, ISSN  1469-1825 , pp. 583-591 , doi : 10.1017 / S0140525X00027412 ( cambridge.org [accessed December 31, 2016]).
21. ↑ Niklas Törneke: Frame of reference theory: An introduction . 1st edition. Junfermann Verlag, Paderborn 2012, ISBN 978-3-87387-791-7 , p. 80 .
22. ↑ Roger Schnaitter: Private Causes . In: Behaviorism . tape 6 , no. 1 , 1978, p. 1-12 .
23. ↑ Niklas Törneke: Frame of reference theory: An introduction . 1st edition. Junfermann Verlag, Paderborn 2012, ISBN 978-3-87387-791-7 , p. 40 .
24. ↑ K. Matthew Lattal: Pavlovian conditioning. In: Gregory J. Madden (Ed.): APA handbook of behavior analysis: Methods and principles. 1st edition. tape 1 . American Psychological Association, Washington, DC 2013, ISBN 978-1-4338-1112-8 , pp. 283-306 , doi : 10.1037 / 13937-013 .
25. ↑ CB Ferster, BF Skinner: Schedules of reinforcement . 1st edition. Appleton-Century-Crofts, New York 1957, p. 10 .
26. ↑ James E. Mazur: Learning and Behavior . 6th edition. Pearson Studium, Hallbergmoos 2006, ISBN 3-8273-7218-6 , p. 548 .
27. ↑ ^{a b} James E. Mazur: Learning and Behavior . 6th edition. Pearson Studium, Hallbergmoos 2006, ISBN 3-8273-7218-6 , p. 201 .
28. ↑ Edmund Fantino, Stephanie Stolarz-Fantino: The logic and illogic of human reasoning. In: Gregory J. Madden (Ed.): APA handbook of behavior analysis: Methods and principles . 1st edition. tape 1 . American Psychological Association, Washington, DC 2013, ISBN 978-1-4338-1112-8 , pp. 439-461 , doi : 10.1037 / 13937-019 .
29. ^ A. Charles Catania, Eliot Shimoff, Byron A. Matthews: An Experimental Analysis of Rule-Governed Behavior . In: Steven C. Hayes (Ed.): Rule-Governed Behavior . 1st edition. Springer US, New York 1989, pp. 119-150 .
30. ↑ Niklas Törneke: Frame of reference theory: An introduction . 1st edition. Junfermann Verlag, Paderborn 2012, ISBN 978-3-87387-791-7 , p. 41 .
31. ^ ^{A b} Mecca Chiesa: Radical behaviorism: The philosophy and the science . 1st edition. Authors Cooperative, Boston 1994, ISBN 0-9623311-4-7 , p. 71 .
32. ^ ^{A b} Michael Perone, Daniel E. Hursh: Single-case experimental designs . In: Gregory J. Madden (Ed.): APA handbook of behavior analysis: Methods and principles . 1st edition. tape 1 . American Psychological Association, Washington, DC 2013, ISBN 978-1-4338-1112-8 , pp. 107-126 .
33. ↑ Christoph Bördlein: Introduction to behavior analysis . 1st edition. Alibri, Aschaffenburg 2015, ISBN 978-3-86569-232-0 , pp. 59 .
34. ↑ Neville M. Blampied: Single-case research designs and the scientist-practitioner ideal in Applied Psychology. In: Gregory J. Madden (Ed.): APA handbook of behavior analysis: Methods and principles . 1st edition. tape 1 . American Psychological Association, Washington, DC 2013, ISBN 978-1-4338-1112-8 , pp. 177-197 , doi : 10.1037 / 13937-008 .
35. ^ ^{A b} Iver H. Iversen: Single-case research methods: An overview. In: Gregory J. Madden (Ed.): APA handbook of behavior analysis: Methods and principles . 1st edition. tape 1 . American Psychological Association, Washington, DC 2013, ISBN 978-1-4338-1112-8 , pp. 3-32 , doi : 10.1037 / 13937-001 .
36. ↑ ^{a b} Anita Jain, Ralf Spieß: Experimental plans for individual experimental research . In: Empirical Special Education . tape 4 , 2012, ISSN  1869-4934 , p. 211-245 .
37. ^ Alan Kazdin: Single-case research designs: Methods for clinical and applied settings . 2nd Edition. Oxford University Press, New York 2011, ISBN 978-0-19-534188-1 , pp. 50 .
38. ↑ Christoph Bördlein: Introduction to behavior analysis . 1st edition. Alibri, Aschaffenburg 2015, ISBN 978-3-86569-232-0 , pp. 45 .
39. ^ S. Watanabe, J. Sakamoto, M. Wakita: Pigeons' discrimination of paintings by Monet and Picasso . In: Journal of the Experimental Analysis of Behavior . tape 63 , no. 2 , 1995, ISSN  0022-5002 , pp. 165–174 , doi : 10.1901 / jeab.1995.63-165 , PMID 16812755 , PMC 1334394 (free full text).
40. ↑ Emiko Uchino, Shigeru Watanabe: Self-recognition in pigeons revisited . In: Journal of the Experimental Analysis of Behavior . tape 102 , no. 3 , 2014, ISSN  1938-3711 , p. 327-334 , doi : 10.1002 / jeab.112 , PMID 25307108 .
41. ^ Richard J. Herrnstein: The Matching Law: Papers in Psychology and Economics . Ed .: Howard Rachlin, David I. Laibson . 1st edition. Russell Sage Foundation Books at Harvard University Press, Cambridge (Massachusetts) 2000, ISBN 0-674-00177-X , pp. 10 .
42. ↑ Christoph Bördlein: Introduction to behavior analysis . 1st edition. Alibri, Aschaffenburg 2015, ISBN 978-3-86569-232-0 , pp. 92 .
43. ^ Zachary J. Zimmermann, Alan Poling: The discipline of behavioral pharmacology. In: Behavior Analysis: Research and Practice . tape 16 , no. 4 , 2016, ISSN  2372-9414 , p. 156-158 , doi : 10.1037 / bar0000057 .
44. ^ Gail Winger, James H. Woods: Behavioral pharmacology. In: Gregory J. Madden (Ed.): APA handbook of behavior analysis: Methods and principles . 1st edition. tape 1 . American Psychological Association, Washington, DC 2013, ISBN 978-1-4338-1112-8 , pp. 547-567 , doi : 10.1037 / 13937-023 .
45. ↑ Mark Galizio: contingency-shaped and rule-governed behavior: instructional control of human loss avoidance . In: Journal of the Experimental Analysis of Behavior . tape 31 , no. 1 , January 1, 1979, p. 53-70 , doi : 10.1901 / jeab.1979.31-53 , PMID 16812123 , PMC 1332789 (free full text).
46. ^ Martha Pelaez: Dimensions of Rules and Their Correspondence to Rule-Governed Behavior . In: European Journal of Behavior Analysis . tape 14 , no. 2 , December 1, 2013, p. 259-270 , doi : 10.1080 / 15021149.2013.11434459 .