CHANGING PROBLEM BEHAVIOR: A SYSTEMATIC AND COMPREHENSIVE APPROACH TO BEHAVIOR CHANGE PROJECT MANAGEMENT

(2008).

CHAPTER 1: REVIEW OF THE

BASIC PRINCIPLES OF

BEHAVIOR

Introduction

What follows is a cursory outline of the basic principles of behavior. This will provide a foundation for the rest of the material. Although the discussion is cursory, I will avoid excessive simplification so that the brief treatment may prove useful even for those familiar with the principles of behavior.

Behavior analysts are concerned with describing and explaining the relationship between behavior and the environment. The environment causes behavior. Our goal is to identify the principles that govern this functional relationship and, specifically, exactly what independent variables in the environment are causing the behavior in any given case. This allows us to make the necessary adjustments to the environment and thereby control the behavior.

Behavior and its cause

Behavior is generally anything that an individual does in response to their environment that can be measured. A response is a particular instance of a behavior. Conditioning (also referred to as learning) is a change in behavior due to experience (exposure to stimuli that cause small-scale physical changes to the body that behaves), and a stimulus is anything that can potentially control behavior (Chance, 2009; Fraley, 2008, p. 64).

There are two types of behavior: respondent behaviors and operant behaviors. Respondent behaviors are reflexive behaviors, which are automatically elicited by an antecedent stimulus and are relatively insensitive to consequences influencing their future likelihood. They include such behaviors as blinking when a puff of air impacts the eyeball, salivating when food is placed in the mouth, and the release of various chemicals into the bloodstream (from glands) that we refer to as emotional behaviors. Respondent behaviors involve smooth muscles, glands and neurons. Operant behaviors are also evoked by antecedents, but they do not occur reflexively; rather, the history of consequences for that behavior influences the likelihood of the behavior being evoked on subsequent occasions. These behaviors involve skeletal muscles, usually large muscle groups. Examples of operant behaviors are the drinking behaviors of reaching for and grasping your fingers around a glass and raising it to your mouth; the walking behaviors of moving one foot in front of the other and shifting weight to move forward; and verbal behaviors, which involve everything you say. Thinking is simply covert verbal behaviors (which are verifiable only to the individual performing the behavior).

We use a contingency or contingency analysis to describe the relationship between the behavior and the environmental stimuli causing it. Respondent behaviors involve an antecedent–behavior contingency, often referred to as a stimulus–response contingency. Operant behaviors involve an antecedent– behavior–consequence contingency, often referred to as a stimulus–response– stimulus contingency.¹

The environment causes behavior by impacting upon and making small-scale changes to the structure of the organism (i.e., stimulation). When we refer to a cause, we are referring to the relationship between two things; one depends on the other in order for it to occur. In science, the thing that is being explained or accounted for (the thing being caused) is called the dependent variable, and the thing that causes it is called the independent variable. Science examines the causal relationship between dependent and independent variables. In the natural science of behavior, the dependent variable is the behavior under consideration, and the independent variable is the environmental stimulation that is necessary and sufficient to produce the behavior (Vargas, 2009, p. 20).

With regard to respondent behaviors, the cause is the antecedent stimulus that elicits the behavior. How does this stimulus cause the respondent behavior? Some respondent behaviors are unconditioned (reflexes that evolved via natural selection), while others become conditioned through pairing of an unconditioned stimulus with a stimulus that did not previously elicit the behavior. With regard to operant behaviors, the cause is also the antecedent stimulus that evokes the behavior. And how does the antecedent stimulus cause the operant behavior to occur? Antecedent stimulation causes an operant behavior because of a history of consequences (specifically, reinforcement) generated by that behavior in the presence of the antecedent stimulus. The consequences are not the cause of a specific response, but rather they influence the future likelihood that the antecedent stimulus will cause the behavior.

Behavior is fully caused by the environment, even though there may be multiple causal variables impacting on whether or not the behavior will result (Fraley, 2008). The principles of behavior examined by the natural science of behavior identify these various causal relationships and how they are established, and provide the foundation for the behavior analytic approach to changing behavior (a technology of behavior). The objective of a behavior change project is to change behavior as efficiently and effectively as possible. First, we identify the actual cause of the specific, observable, measurable behavior in question, and then we use the empirically derived principles of behavior and intervention strategies to establish and achieve specific behavior objectives.

Some say that neurological and physiological processes—the biology of the organism—are the cause of behavior, and others say that behavior is caused by the evolution of the species. The important distinction here is the level of analysis appropriate to the objective. While biology does indeed play a mediating role in behavior and may be one of the more proximate causes, the environmental stimulation that causes these biological processes is the most appropriate level of analysis for changing the behavior of individuals (Fraley, 2008; Vargas, 2009). The level of analysis useful to a technology of behavior change is the functional relationship between the environment and the behavior, and not the mediating biological processes or the ultimate evolutionary processes that resulted in certain anatomical structures. These are topics of interest in certain fields of study, but not primarily for the technologist of behavior.

Operant conditioning

Contingencies

A useful way of conceptualizing the relationship between behavior and the environment is using the three-term contingency. (Contingency refers to the functional or causal relationship between the behavior and the environment.) The three terms in the three-term contingency are the antecedent stimuli, the behavior and the postcedent stimuli. Antecedents and consequences constitute the environmental stimulation that influences the behavior.

The antecedent is what comes immediately before the behavior. Once we have confirmed the functional relationship (that is, that the stimulus actually does cause or reliably control the behavior), we refer to the antecedent stimulus as the discriminative stimulus (SD). There are often numerous stimuli sharing control over behavior, even though we normally simplify the process by identifying the stimulus that exerts the most control.

The term behavior is general and may represent any measurable action an individual exhibits. Particular instances of a behavior are called responses. We avoid speculating about redundant or fictitious constructs that cannot be observed, measured and verified, such as the so-called mind.

The postcedent stimuli involve everything that comes immediately after the behavior. Many things occur after a behavior, and only some of them are functionally related to the behavior—that is, influencing its likelihood on subsequent occasions. Once a functional relationship has been demonstrated, the postcedent stimulus is called a consequence.

The three-term contingency is made up of two two-term contingencies, which together (usually) describe the entire functional relationship between the environment and the behavior. The first two-term contingency is the SD– behavior relationship. The second two-term contingency is the behavior– consequence relationship. These two contingencies are functionally and integrally related. The behavior–consequence contingency functions to strengthen or weaken the SD–behavior relationship or, put another way, to strengthen or weaken the evocative capacity of the SD. Conditioning creates changes to the structure of the body of the learner via stimulation from the environment, which change the evocative capacity of the SD on subsequent presentations.

By determining the three-term contingency that operates in any particular response, we describe the relationship between the behavior and its environment and we therefore explain the cause of the behavior (identifying the dependent and independent variables and their relationship). The principles of behavior discussed below elaborate how these relationships operate.

Below is a representation of the three-term contingency, first in general form and then after the functional relationship has been established for a specific instance of a behavior:

Antecedent → Behavior → Postcedent

SD → Response → Consequence

Sometimes, a four-term contingency is appropriate. When another variable becomes important in controlling behavior, it can be advantageous to include a fourth term in the description of a contingency. For our purposes, including an initial term to account for any function-altering variables that participate in controlling the behavior can sometimes be useful. This is not necessary in a contingency analysis if the three-term contingency provides adequate predictive and explanatory power. However, when another function-altering variable exerts significant control over the behavior of concern and modification of it may be useful in changing the behavior, inclusion of a fourth term becomes useful.

Function-altering stimuli (often broken down into motivating or establishing operations, setting events, sensitization, habituation, potentiation, etc.) will be discussed in more detail below. These function-altering stimuli are distinct from the SD that evokes the behavior; they are stimuli or conditions that provide a context, or additional stimulation that changes the evocative capacity of the SD. Examples are the presence of some other distinct stimulus, conditions of deprivation or satiation of the reinforcer, or medical conditions, such as pain. Such stimuli are often ongoing, rather than presented discretely immediately before the behavior. These context-setting stimuli result in the SD becoming more or less likely to evoke the behavior.

Below is a depiction of a four-term contingency:

SFA → SD → Response → Consequence

More complex multiple-term contingency analyses exist, but these three- and four-term contingencies will suffice for now. Interested readers are directed to Fraley (2008) for discussion of multiple-term contingencies and function-altering stimuli.

Law of effect

The law of effect is the perfect place to start in outlining the principles of behavior. The law of effect simply states that the likelihood of an operant behavior is a function of the consequences it has generated in previous performances. There are, basically, two kinds of consequences: those that produce a satisfying state of affairs and those that produce an annoying state of affairs (in the terminology common when the law of effect was formulated). If a behavior produces a satisfying outcome, that behavior will be more likely again on subsequent occasions. If a behavior produces an annoying outcome, that behavior will be less likely in the future. We no longer define the principles of behavior in terms of pleasant or annoying but simply by whether the consequence does, in fact, increase or decrease the strength of behavior on subsequent occasions. The law of effect is the foundation principle for operant conditioning. If an animal is doing something, it is surely because, in the past, that behavior has contacted a reinforcer. It is getting them something or getting them out of something. It is why all of us do anything we do, operantwise.

We often say that consequences strengthen or weaken the behavior. This is fine when we are trying to simplify the process, but it is not as complete as saying that consequences provide the conditioning history that creates changes within the animal that increase or decrease the probability of the SD evoking the behavior on subsequent occasions of its presentation. To elaborate, the consequence of a behavior does not affect the behavior that preceded it (a teleological fallacy), even though we often say that the animal does such and such in order to get the treats or the consequence strengthens the behavior or the animal does what works. The future does not cause the past. The consequence only changes the future probability of the behavior on subsequent exposures to the SD. And it does not strengthen the behavior directly but rather causes changes in the animal’s body (stimulation) that thereby change the probability of the behavior being performed on subsequent occasions (conditioning)—learning occurs physically within the subject.

We will discuss five kinds of consequences, three that weaken behavior and two that strengthen behavior. The two that strengthen behavior are referred to as reinforcement; two of the three that weaken it are referred to as punishment. These four principles involve changes to the postcedent environment. They are different from the fifth kind, extinction, which involves an absence of postcedent environmental change for previously reinforced behaviors—in other words, withholding reinforcement. In reinforcement and punishment procedures, you can present a stimulus following the behavior (a positive consequence), or remove a stimulus following the behavior (a negative consequence). This means that you can have positive reinforcement and negative reinforcement, and positive punishment and negative punishment.

Figure 1 provides a flow chart that illustrates how to analyze any particular operant conditioning instance and identify the principles operating on the behavior in question.

Figure 1. Flow chart showing basic principles of behavior.

The easiest way to conceptualize the types of operant conditioning and their relationships with one another is in the contingency diagram depicted in Figure 2. Many readers will not be familiar with this diagram and will be more familiar with a table showing the four quadrants of operant conditioning. The main problem with the quadrants table is that it addresses only changes in behavior occurring due to postcedent changes and ignores extinction, the fifth basic principle of operant conditioning. By eliminating the table format and using a dimensional format, we can account for extinction diagrammatically. As Fraley (2008, p. 400) explains, [w]hen the rate goes down on future occasions of any behavior that continues to yield no change in the environment (no consequence), that behavior is extinguishing. Extinction is different from the other four principles. The other four principles occupy the corners of the diagram because they involve changes to the postcedent environment (presentation or withdrawal of stimulation, the left–right axis in the figure) as well as an increase or decrease in the strength of the behavior (the up–down axis). Extinction, on the other hand, involves no functional postcedent change in the environment from the antecedent environment for a behavior with a history of reinforcement. It therefore exists only directly on the vertical axis line; because the behavior decreases in rate or frequency, extinction is found at the bottom of the figure. Don’t be confused into thinking that extinction results from punishment (which would require a left–right axis position in the figure). It does