Stimulus Equivalence Research Paper

This sample Stimulus Equivalence Research Paper is published for educational and informational purposes only. If you need help writing your assignment, please use our research paper writing service and buy a paper on any topic at affordable price. Also check our tips on how to write a research paper, see the lists of psychology research paper topics, and browse research paper examples.

In stimulus equivalence, untaught abilities arise spontaneously out of a few that are learned through direct experience. Students of psychology often wonder where the most creative and flexible aspects of thought come from; stimulus equivalence may provide a partial explanation. Until recently, scientists had a limited understanding of the experiences that create stimulus equivalence. Contemporary research on the building blocks of stimulus equivalence provides insights into some complex everyday problems, including that of how to teach for maximum benefit. This research-paper defines stimulus equivalence, provides some tips for understanding the rather technical methods that are used to study it, and illustrates some applications of the insights that are derived from the research. The chapter also shows how stimulus equivalence is related to some well-known psychological phenomena such as operant learning, concept learning, and knowledge networks in long-term memory.

Introduction

Among our varied psychological gifts, perhaps most valuable is the ability to “go beyond the information given”— that is, to derive more knowledge than has been learned through direct experience. Stimulus equivalence describes such an ability.

Starting about a century ago, John Dewey (1933), whom many consider a father of contemporary educational thought, placed “going beyond the information given” at the center of his views on education. In mastering a new academic subject, he said, we begin by memorizing new facts, but our potential intellectual power goes far beyond that. The goal of education, in fact, is to promote “reflective thinking” that integrates and extends what had been learned through memorization. In a now-famous example of “relective thinking,” Dewey noted that a student who learns that All men are mortal and that Socrates is a man should be able to derive—without any further instruction— that Socrates is mortal.

Dewey’s syllogism captures the spirit of stimulus equivalence, a derivation of untaught relations from what has been learned through direct experience. A “relation” can be thought of as a psychological connection between events. Stimulus equivalence exists when we treat relate dissimilar stimuli and treat them as functionally (psychologically) interchangeable. As will be shown later, to scientists “responding similarly” is defined in a very specific way. In general terms, however, think about how “responding similarly” might work in an example that closely parallels Dewey’s. Imagine three stimuli: the spoken word “dog” (A), a Doberman pinscher (B), and the printed word dog (C). To our senses, these stimuli have little in common, but with the right past experiences we react to them in much the same way. For instance, a glimpse of a Doberman, a printed sign stating Beware of the dog, and the spoken warning, “Watch out for that dog!” all can prompt the same precautionary measures.

Figure 39.1 A simple stimulus equivalence class incorporating three stimuli. Relations that were learned by direct experience are shown as solid arrows. Those that emerged, untaught, are shown as dashed arrows.

Experience sometimes explicitly teaches us to treat nonidentical stimuli as equivalent—think of the vocabulary exercises involved in learning a second language. “Reflective thinking,” however, implies emergent equivalence. Assume that a child has been explicitly taught to associate the spoken word “dog” (A) with a photograph of a Doberman (B), a relation that can be expressed as A—’B (the arrow notation implies that when A, “dog,” is present, B prompts some response, such as pointing to a picture of a Doberman [rather than, say, a wombat]). The child also has learned to associate “dog” with the printed word dog (A—C). These explicitly taught abilities are shown as solid lines in Figure 39.1. With these abilities in place, other abilities can emerge without further teaching. Most notably, the child should now relate the picture of the Doberman to the printed word dog (A— C), even though these two stimuli have never been experienced together.

Tracy Zinn (2002), a graduate student, used stimulus equivalence to address a learning problem that her professor encountered while teaching a college course on drugs and behavior. Students often became confused because commercial medications have both a brand name and a generic name (see Chapter 17, Drugs and Behavior). For example, Valium® is a brand of the generic antianxiety drug diazepam, and Thorazine® is a brand of the generic antipsychotic drug chlorpromazine. Because many drug names are difficult to pronounce, some students also had trouble connecting spoken drug names with their written equivalents. Overall, students spent so much time trying to learn drug names that they were unable to focus on more interesting features of psychoactive drugs. Zinn showed that instruction based on stimulus equivalence can help people learn drug names with less effort.

Zinn’s students worked individually on a computerized instructional module that taught selected relations between trade and generic names of 32 medications in four different drug classes. For each drug, a Stimulus Equivalence Group was taught the relations shown as solid arrows in Figure 39.2 (left side). They learned to recognize written trade and generic names when given the spoken versions of those names, and also to match the written trade name to the spoken generic name. Afterward, the students were tested on three emergent relations among these names (dashed arrows in Figure 39.2).

A “Complete Instruction” Group worked on a different computer module, one that explicitly taught all of the relations shown in Figure 39.2 (left panel), before proceeding to the same posttest as the Stimulus Equivalence Group. Not surprisingly, this group took about twice as long to complete their instruction. As shown in the right panel of Figure 39.2, both groups mastered the three relations that were taught explicitly to the Stimulus Equivalence Group. The groups also did equally well on the three relations that the Stimulus Equivalence Group was not taught.1 Thus, this study showed how stimulus equivalence can contribute to instructional efficiency. Both groups succeeded, but for the Stimulus Equivalence Group, “reflective thinking” substituted for a large portion of the instructional effort required to build mastery in the other group.

An “Engineering” Challenge

As Dewey’s syllogism suggests, philosophers and psychologists have long recognized that stimulus equivalence can occur. Many theorists have not been very clear about what experiences are needed to promote the “reflective thinking” that unleashes our full intellectual power. A major contribution of stimulus equivalence research has been to specify some of the conditions under which new abilities reliably emerge—untaught—from those learned through direct experience. Among other things, such knowledge allows “reflective thinking” to be created at will, as the following example illustrates.

Formal Definition of Stimulus Equivalence

Untaught abilities like those described above can arise spontaneously once an individual has mastered a few overlapping relations among stimuli. Up to three kinds of untaught abilities (dashed lines in Figure 39.1) can emerge. Scientists label these abilities using terms borrowed from mathematical set theory (for the logic behind this, see Sidman, 1994), and when all three are present, a stimulus equivalence class is said to exist.

Figure 39.2 How some students efficiently mastered the relations among drug names. Left side: Relations acquired through direct experience are shown as solid arrows. Those that emerged, untaught, are shown as dashed arrows. Right side: Students taught using stimulus equivalence learned as well as students taught more traditionally, but did so after only about half of the effort. See the text for further details.

SOURCE: Based on Zinn, 2002.

The emergent ability that best captures the spirit of “reflective thinking” is called transitivity. It involves relating stimuli that, although never previously paired in direct experience, share a common associate. In Figure 39.1, direct experience has not taught that the printed word dog applies to the Doberman photograph (B—>C), or vice versa (C—>B), but they may be related by virtue of their common connection to the spoken word “dog” (A).

Psychology students often wonder about the essential ingredients of human nature. Quite possibly, transitivity is one of them. Nonhuman animals show several kinds of emergent abilities, but to date it is not clear whether transitivity is within their reach. Marine Biologist Ronald Schusterman and his colleagues (2003) believe that they have seen transitivity in sea lions, although not everyone agrees with their conclusions. This debate addresses important questions about what it means to be human, but for now it should not distract us from a more fundamental point. As John Dewey lamented, without the right learning experiences, even the brightest among us may not show transitivity spontaneously.

Stimulus equivalence also implies the existence of two other kinds of relations that often are taken for granted in everyday experience. Symmetry is the inverse of an explicitly learned relation between stimuli. Having learned that “dog” refers to a Doberman (A— B), an individual may infer that the Doberman is labeled as “dog” (B— A). In reflexivity (circular arrows in Figure 39.1), an individual recognizes a stimulus as itself. Reflexivity is required in a common child’s game in which a picture (e.g., of a Doberman) must be matched to an identical picture among several pictures (Doberman, wombat, dolphin) spread out on a table. Although not illustrated in the figure, reflexivity also is at the core of recognition memory (see Chapter 46, Memory), which would be required in the child’s game if the first Doberman picture were put away before presenting the choices on the table.

At first blush, symmetry and reflexivity seem uninteresting because typical adults so often show these abilities effortlessly, and not all stimulus equivalence research studies focus on them (e.g., note that the study on drug names tested only for transitive and selected symmetrical relations). Yet symmetry and reflexivity are part of the definition of stimulus equivalence for two reasons. One reason is logical: In mathematical set theory, which provides our vocabulary for talking about the relations involved in “reflective thinking” (for more about this topic, see Sidman, 1994), equivalence exists among members of a group of stimuli only if reflexivity and symmetry also exist. The other reason is practical: These abilities do not always emerge spontaneously. For example, in the case of symmetry, an English-speaking student of the German language may grasp that dog means hund before mastering the translation in reverse. In the case of reflexivity, consider a young medical student who, having felt an abnormal mass in cancerous breast tissue for the first time, now struggles to identify an identical mass in a different breast. As these examples suggest, for typical adults, symmetry and reflexivity may be most problematic when unfamiliar stimuli are involved. These relations also can be problematic for very young children, persons with intellectual disabilities, and nonhumans. Such cases create enough uncertainty that many stimulus equivalence studies, even with typical adult participants, test for symmetry and reflexivity, just to be sure they occurred.

Inheritance of Function

The stimuli in an equivalence class may be thought of as psychologically interchangeable. Before coming together in an equivalence class, some stimuli may affect us in very specific ways (e.g., they cause fear) while others are unfamiliar and thus essentially meaningless. When an equivalence class forms, the stimuli may each inherit any psychological functions served by other class members. Two examples to illustrate this inheritance of function follow.

Psychologist Dermot Barnes-Holmes and his colleagues have shown how inheritance of function might influence consumer brand-name product preferences (D. Barnes-Holmes, Keane, Y. Barnes-Holmes, & Smeets, 2000). In an early part of the experiment, two equivalence classes formed. One paired nonsense syllables (e.g., VEK, YOF) with words that arouse positive emotions (e.g., “holidays”); the other paired different nonsense syllables with words that arouse negative emotions (e.g., “cancer”). Through inheritance of function, the nonsense syllables, originally psychologically neutral, now would be expected to elicit emotions. In a later part of the experiment, the nonsense syllables were paired with brand names of two carbonated soft drinks. Thus, brand names were related to emotional stimuli, but only indirectly. Finally, participants drank from bottles with different brand-name labels but identical contents. If the flavor of the drinks were all that mattered, there would be no reason to prefer either brand. Instead, the “consumers” clearly preferred the brand that was indirectly associated with pleasant words, an outcome that makes sense in terms of inheritance of emotional functions.

Figure 39.3      An experiment illustrating the inheritance of learned fear through an equivalence class. See the text for details.

Many kinds of psychological functions are known to spread through equivalence classes (Sidman, 1994), but to illustrate how inheritance of function might be clinically relevant, for our second example let’s stick with the triggering of emotions. Several clinical disorders incorporate learned fears of harmless stimuli. What if one of the stimuli in the equivalence class of Figure 39.1 became a conditioned elicitor of fear? A team led by clinical psychologist Michael Dougher (1998) has found that learned fear can spread through equivalence classes. To get a sense of how their experiments worked, look at Figure 39.3. The three stimuli in the top left box, including the spoken word “dog,” probably do not elicit fear under normal circumstances. Imagine, however, that a child’s mother tells a terrifying story (indicated by a lightning bolt) about the time she was attacked by a neighbor’s dog. The story elicits fear in the child, and, through a possible case of rapid classical conditioning (see Chapter 33, Classical Conditioning), the spoken word “dog” now does as well. Through inheritance of function (middle box), pictures of dogs and the printed word dog should now elicit fear, too, even though they were not part of the conditioning episode. This effect—which could also encompass things related to dogs like leashes and fire hydrants—may help to explain why the learned fear of a single object often blossoms into fear of many things. Finally, if the child’s fears are clinically problematic, he may receive an exposure-based therapy like systematic desensitization (see Chapter 80), which is thought to harness extinction of classical conditioning. Fear-eliminating benefits should then spread through the equivalence class (right box), even if most of the stimuli are not directly involved in the therapy.

Generativity

An experience is generative if it spawns novel abilities. The essential feature of stimulus equivalence is that many abilities emerge spontaneously after direct experience learning only a few component relations. In the example of Figure 39.1, two abilities were taught (relating A to B, and relating A to C). As many as seven others emerged, including two transitive relations (B—’C, C—B), two symmetrical relations (B—A, C— A), and three reflexive   relations   (A— A, B— B, C— C). Critically, these emergent abilities arise from experience with a properly structured “curriculum” that involves interconnected relations among stimuli. An individual who learns disconnected relations (e.g., “dog”— Doberman photo and smiley-face icon— printed word happy) can expect no emergent benefits.

Figure 39.4  Left box: Each new stimulus added to a class greatly expands the number of emergent relations. Both boxes: Classes also can expand by merging with other classes. In both cases, see the text for explanation.

Boosting Class Size

The number of possible emergent abilities increases with the number of stimuli that are imbedded in interconnected relations. To illustrate, the left box of Figure 39.4 shows the implications for the person whose experience is summarized in Figure 39.1, that of learning that the printed word canine implies dog (for now, ignore the right box and the arrow connecting the two boxes). Learning this one ability spawns as many as six new emergent ones: four transitive, one symmetrical, and one reflexive (these relations are designated by stars in the left box). Several different approaches exist for building connections between relations (see Green & Saunders, 1998, for details). For many of these relations, the ratio of directly learned to acquired relations is N-1 to N2, with N representing the number of stimuli in the class. Thus, in a three-member class, as per Figure 39.1, the ratio is 2 directly learned relations to 9 acquired (7 emergent). In a four-member class, as per Figure 39.4, the ratio is 3 to 16 (13 emergent). In a five-member class, the ratio is 4 to 25 (21 emergent), and so on.

Another factor that contributes to generativity is that two classes sharing a common member may merge into one bigger class. As illustrated in Figure 39.4 (right box), imagine learning that man’s best friend ◊ likes fire hydrants, and man’s best friend ◊ likes bones. This experience should create an equivalence class that includes the transitive relation between likes bones and likes fire hydrants. Now, as shown by the arrow spanning the two boxes, imagine that you subsequently learn that dog ◊ man’s best friend. This experience places man’s best friend into the dog class, and dog into the class containing man’s best friend. As a result, every member of the dog class (A, B, C, D) should be readily associable with every member of the class containing man’s best friend (E, F, G), and vice versa. Many new relations emerge between stimuli that have never been experienced together (such as “dog” ◊ likes fire hydrants and likes bones ◊ canine). In this case, teaching a single new relation yields a total of up to 22 new emergent transitive relations.

Connections

Stimulus Generalization And Cognitive Transfer

As will be described later, stimulus equivalence is thought to arise out of operant learning, but it is different from the operant learning concepts you probably read about in your introductory psychology textbook (see Chapter 36, Operant Conditioning). The purpose of this section is to clear up a common confusion about how stimulus equivalence is related to operant learning.

Brief Review of Operant Learning

Operant behavior is altered by its consequences, or the events that follow behavior. Among consequences, reinforcers increase the probability that a given behavior will be repeated, and punishers decrease that probability. For example, a young student’s first days in public school are full of these consequences. Some actions (e.g., sitting quietly and listening) are reinforced (e.g., by teacher praise) and over time they begin to increase in frequency. Others (e.g., eating chalk) are punished and begin to disappear.

Operant learning is situation-bound. What we learn through reinforcement and punishment tends to apply mainly to the situations in which the original learning occurred (Dinsmoor, 1995). Thus, a student can learn appropriate classroom behavior from consequences, but this skill set is expressed mainly while in school. Students do not suddenly whip out books and begin taking notes in a supermarket, restaurant, or automobile dealership. Operant behavior becomes situation-specific when the consequences that fuel it also are situation-specific. For example, academic behaviors are reinforced in some settings but not others. Behaviors like taking notes and sitting quietly with books have beneficial consequences in the classroom, but not in a restaurant. The term for a situation correlated with reinforcement of a given behavior is discriminative stimulus. By uniting the demands of the present moment with the benefits of past experience, a discriminative stimulus creates a situation-specific increase in the probability of acting a certain way (Dinsmoor, 1995). For academic behaviors, the classroom is a discriminative stimulus, but a restaurant is not.

Stimulus Generalization

Once discriminative stimuli exist, the benefits of past experience can spread to new situations through stimulus generalization. That is, operant learning spontaneously transfers to new situations that are perceptually similar to the discriminative stimulus (Dinsmoor, 1995). This process is how sitting quietly and taking notes can occur, without special training, the first time a student attends a new class or perhaps, when older, a professional conference.

Analogously, cognitive psychologists have long been interested in the study of transfer, which they define as the application of knowledge to novel contexts (Bransford, Brown, & Cocking, 2000). For instance, imagine that when quite young you learned to make change during a sale while working in your parents’ business. You would show transfer if you could now also solve mathematics problems not involving money. This type of transfer shares two important characteristics with operant generalization: It involves the application of a learned ability to a new situation, and it is most likely to occur when there are close similarities between the original learning situation and the transfer context.

Different Processes

Stimulus generalization and cognitive transfer are special cases in which inheritance of function is mediated by perceptual similarities between the original learning situation and new stimuli. A conference hall, for instance, shares many features with a classroom (orderly rows of chairs, podium or desk at the front, etc.). When first encountered, it is “treated as the same” as a classroom because it is, in fact, similar to the classroom. By contrast, when first encountered a neighborhood bar is “treated differently” from a classroom because it is perceptually different from the classroom.

Importantly, stimulus equivalence is not stimulus generalization (transfer). Stimulus equivalence does not require that stimuli share perceptual features to be “treated as the same.” Recalling Figure 39.1, the auditory waveforms of the spoken word “dog,” the ink marks of the written word “dog,” and the color patterns of the Doberman photograph would not be mistaken for one another, even by an alien visitor to our planet. They “go together” only insofar as experience (e.g., the arrows of Figure 39.1) teaches that they do.

A Productive Collaboration

Although stimulus equivalence and generalization (transfer) are different, these processes may combine to dramatically increase the size of equivalence classes (Fields & Reeve, 2000). Note, for instance, that although no two photographs of a Doberman are identical, most Doberman photographs look pretty much like any other. Doberman photographs also share perceptual features with some drawings of dogs, some real dogs, and some toy stuffed dogs. Practically speaking, in Figure 39.1 the B stimulus can represent not just a single Doberman photograph but a vast range of stimuli with shared properties. The equivalence class in Figure 39.1, therefore, incorporates not three stimuli but perhaps many hundreds or thousands. To be clear, overlapping stimulus relations place the first Doberman photograph into this class, whereas stimulus generalization then recruits many similar stimuli into it.

Semantic Generalization and Knowledge Networks

Stimulus equivalence may well remind you of other concepts you have encountered in your study of psychology because psychologists have long assumed that thinking is organized around networks of interconnected knowledge or ideas. Here are just three examples. First, the brain contains networks of synapses connecting neurons (see Chapter 3, Biological Psychology), which has prompted speculation that complex thought arises from these networks. Second, neural networks are the inspiration for computer programs that can simulate complex psychological abilities. The software building blocks of these programs are intended to mimic interconnected networks of neurons that fire together under certain circumstances (see Chapter 52, Artificial Intelligence). Third, some theorists believe that the contents of long-term memory are organized into semantic networks in which bits of knowledge are linked through shared meaning rather than physical similarity (see Chapter 46). All of these network theories share a common theme with stimulus equivalence: Individual parts of a knowledge network combine to make emergent abilities possible. In all of these cases, the whole of thought is seen as greater than the sum of its parts.

One interesting phenomenon that network theories anticipate is “semantic generalization,” in which learning transfers to new situations on the basis of shared meaning rather than perceptual similarity. Consider a phobia patient with a crippling fear of spiders. Stimulus generalization may lead him to fear tomatoes because the cluster of slender leaves atop this fruit resembles a spider’s leg. Later, through “semantic generalization,” he may come to fear a variety of fruits and vegetables, but not because they remind him of spiders. Rather, through experience, these foods are understood similarly to tomatoes (for instance, they all come from plants, bear seeds, and make for healthy eating).

Knowing about stimulus equivalence can advance our understanding of “semantic generalization” in two ways. One advantage is that we can identify a contradiction in the term used to label this effect. Recall that, in “semantic generalization,” transfer is mediated by shared meaning between the original learning situation and new stimuli. By definition, however, in generalization, transfer is mediated by perceptual similarity. Thus “semantic generalization” is not really generalization. Instead, it may be an example of the inheritance of function that we expect to see in equivalence classes (for instance, for most people who eat plant products, tomatoes probably are in an equivalence class with corn, carrots, and kumquats). Thinking of “semantic generalization” in this way avoids a confusion of terms. A second advantage comes from our ability to meet the “engineering challenge” mentioned earlier in this research-paper. Some network theories do a better job of describing knowledge networks than of showing how to create them. By contrast, we know a great deal about how equivalence classes form, and this knowledge suggests a clear way in which the shared meaning underlying “semantic generalization” could arise through experience.

Concept Learning

Some psychologists believe that knowledge networks are created through concept learning. Speaking loosely, a concept is a collection of things that “belong together,” and concept learning is the process of determining which things “belong together” (Barsalou, 1992). For example, a person who has acquired the concept of chairs not only recalls many specific chairs from past experiences but also recognizes new chairs, never before seen, for what they are. At the root of concept learning is experience, with feedback, of reacting to examples and nonexamples of the concept (similar experience helps to form equivalence classes). Imagine, for example, the experience of an alien visitor to our planet. This visitor might treat some objects as “chair” by trying to sit in them—sometimes with success (the object supports weight comfortably) and sometimes not (the object collapses or feels uncomfortable). The visitor might also react to some objects as “not chair” by continuing to stand—this decision works well in some cases, and in others the visitor is corrected by being told to “have a seat.” Such experience creates two outcomes that are reminiscent of stimulus equivalence. First, various nonidentical examples and ideas coalesce into a kind of knowledge network. Second, a defining property of these knowledge networks is inheritance of function: When we learn something new about one example of a concept, our understanding of others automatically is updated.

There are at least two important differences between stimulus equivalence and concept learning. First, many concepts are defined by physical similarities among stimuli, while the stimuli in an equivalence class are united by shared function (Fields & Reeve, 2000; Zentall, Galizio, & Critchfield, 2002). Second, unlike in equivalence classes, the stimuli in some concepts are not interchangeable (Barsalou, 1992). They are united instead by other kinds of relations. For example, your favorite recliner is one of many chairs (equivalence relation), but it is also a component of your living room suite (part-to-whole relation). Psychologist Thomas Zentall, who has studied concept learning extensively, suggests that we think of equivalence classes as a special type of concept (Zentall et al., 2002).

Methods

A good understanding of stimulus equivalence will require reading beyond this research-paper including primary empirical reports. The procedures used to study stimulus equivalence in the laboratory are rather technical (Green & Saunders, 1998), and as a result, many people find them to be difficult reading. The goal of this section is to explain enough about the methods of stimulus equivalence research to ease the burden of students who seek to explore published reports. Remember, however, that experimental procedures do not define stimulus equivalence; rather, they allow scientists to view the development of equivalence classes more clearly than would be possible in everyday situations. The “father of stimulus equivalence research,” Murray Sidman (1994), has stressed that outside of the laboratory equivalence classes form under very different circumstances than those seen in the laboratory.

Training, Testing, and Experimental Designs

Every stimulus equivalence experiment has at least two parts. During a training phase, overlapping stimulus relations (like those described by solid arrows in Figure 39.1) are explicitly taught, with feedback provided for correct and incorrect responses. Although the relations taught in many studies seem quite simple, sufficient practice is provided to guarantee that they are learned well. In published studies, you’ll see that intelligent people often master these relations easily, but inexplicably, occasionally they require a surprising amount of practice. For example, in an experiment conducted by one of the authors of this research-paper, a college-student participant needed 648 trials to learn three AjB relations involving “simple” geometric stimuli. What is “simple,” apparently, is in the eye of the beholder!

After training concludes, a testing phase follows in which untaught relations (analogous to those defined by dashed arrows in Figure 39.1) are assessed. Laboratory research studies often test for all possible reflexive, symmetrical, and transitive relations. Applied studies may focus only on transitive relations, as in the drug-names study (Zinn, 2002). Note that, although feedback is a critical part of training (see the section on “extrinsic” motivation, below), none is provided during the test, so that “emergent” relations can be shown to be truly untaught. Participants may, however, be confused by this sudden loss of feedback. Thus, in many studies, at the end of training, feedback is gradually eliminated from some trials to prepare participants for the change.

The results of a stimulus equivalence study may be presented at two levels of analysis. First is the general question of how many participants met the definition of stimulus equivalence by demonstrating all of the reflexive, symmetrical, and transitive relations. Second is the more specific question of how many of the participants mastered each potential emergent relation.

Some experiments, like the drug-names study (Zinn, 2002), use familiar group-comparison experimental designs involving many participants. Others, however, merely demonstrate that a particular set of training experiences is sufficient to create equivalence classes. Many of these studies focus on as few as four to six individuals.3 which is likely to raise red flags for students who have been taught in statistics class to distrust small samples. The small N (see Chapter 10, Single-Subject Designs) is used for two reasons. First, when all individuals move seamlessly from random responding to complete mastery, there is little need for inferential statistics to verify the results. Second, the complex performances that participants learn require very detailed analysis, something that is easiest to describe for only a few individuals.

“Artificiality” of Laboratory Research

Stimulus equivalence has been the inspiration for a variety of attempts to teach socially significant abilities. In laboratory research, however, great emphasis is placed on understanding how new equivalence classes form, and the resulting study may not look much like the everyday world. The present section describes a few ways and reasons that laboratory research procedures may be different from the conditions in which equivalence classes form outside the laboratory.

Skill-Building Emphasis

Although people form equivalence classes in everyday experience, most research does not focus on analyzing these preexisting classes. One reason is that, because each individual has a unique history, such classes may be different for different people. For example, you might associate tequila with pleasant travels and social events while your friend associates it with sin and damnation. Another reason is that scientists have difficulty figuring out what everyday experiences, at some undetermined point in the past, created each individual’s equivalence classes. Recalling the “engineering challenge” mentioned earlier, basic research usually seeks to understand general rules that govern how equivalence classes are formed rather than conditions that created a specific class for a specific person, and it does so by building new classes from scratch.

Unusual Stimuli

In any study of learning, it is important to verify that new abilities arise from experience in the study, rather than from a prior learning history. Thus, the stimuli used in basic research may be chosen because they are not meaningfully related to everyday experience. Instead of stimuli like those shown in Figures 39.1 and 39.2, many studies use stimuli that are unlikely to trigger strong associations prior to the start of an experiment (see Figure 39.5).

Structured Procedures

Because equivalence classes are defined by relations among stimuli, researchers use experimental procedures that precisely define these relations. Most typically, stimuli are presented in a match-to-sample procedure, one variant of which is illustrated in Figure 39.5. In the top leftmost panel, a sample stimulus is presented. The participant may be required to point to it, or to position a mouse cursor over it, to show that she is paying attention (top row, second panel), which reveals two or more comparison stimuli that are physically unlike the sample (top row, third panel). The researcher has arbitrarily decided that one of these “goes with” the sample, and the participant’s job is to learn from feedback about this correspondence. In Figure 39.5, an octagon comparison corresponds to the sample zeta (top row); a circle comparison corresponds to the sample phi (middle row); and a triangle comparison corresponds to the sample delta (bottom row). Thus, the procedure teaches, not a simple discrimination in which some stimuli are correct and others incorrect, but rather a conditional discrimination in which the correct choice depends on which sample is present.

“Extrinsic” Motivation

Figure 39.5  The match-to-sample procedure and examples of unfamiliar stimuli, often used in laboratory research on stimulus equivalence.

If the correct stimulus is chosen, feedback indicates that the response is correct (Figure 39.5, top and bottom rows, last panel). Otherwise, corrective feedback follows (middle row, last panel). Feedback not only shapes the subject’s skills; it also provides a form of motivation that is important to completing the experiment. Unlike many everyday stimuli, stimuli like those in Figure 39.5 are not inherently interesting. Because most people care about being correct, feedback gives participants a reason to pay attention and to learn. To enhance motivation further, some studies accompany feedback with money (e.g., two cents for each correct response) or other tangible benefits.

Multiple Classes

All experiments attempt to create two or more equivalence classes simultaneously. Each of the relations shown in Figure 39.5 (zeta-octagon, phi-circle, and delta-triangle) would be part of a different potential equivalence class (e.g., in A-B-C format, the classes might be: zeta-octagon-*, phi-circle–, and delta-triangle–). For economy of presentation, only the training of A j B relations is shown. Creating multiple classes provides a built-in test of replicability—whatever works in creating one class should work in creating the others. Notice, too, that the comparison stimuli are the same in all parts of Figure 39.5. Training for multiple classes allows stimuli that are correct in one relation to be used as incorrect comparisons in others, which, in turn, eliminates some potential confounds that give the impression of participants understanding more than they really do (see Green & Saunders, 1998, for a nice explanation of this problem).

Applications

 Instruction

John Dewey, whose thoughts on “reflective thinking” began this research-paper, also argued,

The value of any fact or theory as bearing on human activity is, in the long run, determined by practical application—that is, by using it for accomplishing some definite purpose. If it works well—if it removes friction, frees activity, economizes effort, makes for richer results—it is valuable. (McLellan & Dewey, 1908, p. 195)

From this perspective, research on stimulus equivalence has shown considerable promise. Because basic research on stimulus equivalence has focused on building novel repertoires from the ground up, its procedures can, with relatively little modification, be employed to build socially relevant repertoires (Stromer, Mackay, & Stoddard, 1992). Applied stimulus equivalence studies always differ from their basic-research counterparts in one primary way: the stimuli are integral to real-world skills.

Because many academic topics require fluency with core equivalence relations, the range of possible stimuli is quite large. Sometimes stimulus equivalence is used to give an extra assist to people with intellectual challenges. In these cases, the relations worth teaching may be quite elementary (Figure 39.1 is a possible example). Sometimes stimulus equivalence is used for instructional efficiency in the teaching of typically developing children. Equivalence-based procedures have been used to teach important academic skills like the rudiments of a second language (e.g., equivalences between English and Spanish words) and of basic mathematics (e.g., equivalences between fractions, decimals, and the quantities they represent). Yet equivalence classes also are imbedded in more sophisticated knowledge. One study, for example, used equivalence-based instruction to teach college students the relations between factored and unfactored algebraic equations and the graphically portrayed functions that the equations represent (Ninness et al., 2005).

When thinking about the instructional potential of stimulus equivalence, it is important not to get too caught up in the detailed procedures of formal experiments. For instance, Zinn’s (2002) drug-names study (see Figure 39.2) nicely illustrates how stimulus equivalence can inform the teaching of fairly complex information, but like all research, it focused as much on documenting an effect scientifically as on enhancing the lives of the participants. Murray Sidman (1994) has noted that although good instruction is carefully structured, it may omit many of the methodological bells and whistles seen in formal experiments. For example, training and testing may be intermingled, detailed explanations may complement the training experiences, emergent relations may be reinforced as soon as they occur, and reflexive and symmetrical relations may not be tested. As in laboratory studies, of course, the critical ingredient is a curriculum of overlapping stimulus relations. To get a sense of how learning experiences can be streamlined from the strict protocols of the laboratory, try teaching yourself a few facts on drugs and behavior by following the procedures described in Table 39.1.

Table 39.1       A streamlined approach to learning facts about psychoactive drugs through stimulus equivalence. Construct a deck of index cards containing the information shown below. Study them by looking at the front, and saying or writing what’s on the back. Shuffle the deck each time you use it, and study until you can run through the deck without error. Then, take the test in the Appendix.

Interpretations Of Pressing Social Problems

Some stimulus equivalence experiments are intended to model processes that are thought to contribute to socially important problems. Recall earlier examples on the development of consumer brand-name preferences and the spread of learned fear. Other experiments have explored the possible role of stimulus equivalence in sexual arousal, fantasy, and deviance (Barnes & Roche, 1997); in how individuals understand social relations (Schusterman, C. R. Kastak, & D. Kastak, 2003); and in social stereotyping, as illustrated by the following research-based example (described by Dixon, Dymond, Rehfeldt, Roche, & Zlomke, 2003). Perhaps you are told that, on average, the people of a country we’ll call Nation X drink more alcohol than people in many other countries (i.e., Nation X I heavy drinkers). Experience also tells you that people who have overindulged in alcohol are not in peak intellectual form (heavy drinkers ◊ dimwitted). Through intuitive inference, you may have low intellectual expectations for people of Nation X (Nation X ◊ dimwitted). This false syllogism may be hard to resist, and illustrates how “reflective thinking” is not always beneficial.

An understanding of equivalence classes has helped theorists to speculate on the origins of many pressing problems like child abuse (Keenan, McGlinchey, Fairhurst, & Dillenberger, 2000), psychological disorders like depression (Dougher, 1998), and prejudice (Dixon et al, 2003). These interpretations are too involved to summarize here, but a brief example may help to illustrate what is involved. Dixon et al. proposed that emergent social stereotyping, similar to what was described in the last paragraph, can explain the hostile attitudes that many people in the United States hold, in the wake of the 2001 terrorist attacks, toward individuals of Middle Eastern descent. A similar process may underlie the hostile attitudes toward the United States of people in underdeveloped countries (including some who become terrorists). Dixon et al. believe that an equivalence-based analysis of these problems has two practical benefits. First, the analysis predicts that some popular educational strategies for reducing prejudice actually will magnify the problem. Research bears out this prediction. Second, the analysis suggests some novel strategies for reducing prejudice and, potentially, terrorism.

Theory

Implications Of Stimulus Equivalence For Theories About Behavior

Stimulus equivalence research arose out of behavioral psychology (see Chapter 2, Psychology in the 20^th Century), which is based on learning processes like operant learning. Psychologist Stephen Hayes, who has led the effort to apply stimulus equivalence concepts to clinical and everyday phenomena, has argued that stimulus equivalence forces two changes in how we look at behavioral psychology. First, behavioral psychology sometimes is criticized as too simple to account for complex human abilities such as language and thought (Barsalou, 1992). Because of its generative potential, however, stimulus equivalence may be one contributor to these abilities.

Second, stimulus equivalence may clarify some specific aspects of theories about behavior. Consider conditioned reinforcers, which become effective consequences through learning (see Chapter 36, Operant Conditioning). Traditionally, this type of learning is thought to be similar to what Russian physiologist Ivan Pavlov observed in his famous studies of conditioned reflexes in dogs. You may recall from your introductory psychology course that Pavlov paired an unimportant sound (tone) with a stimulus (meat) that elicited an unlearned reflex (salivation). With repeated pairings, the tone began to elicit salivation. Analogously, most textbooks will tell you that conditioned reinforcers arise when formerly unimportant stimuli are paired directly with stimuli that already work as reinforcers (e.g., see Dinsmoor, 1995). A smile, for instance, is reinforcing to most people, but not originally to infants. It may become reinforcing to the extent that infants encounter caretaker smiles at the same time as reinforcers such as food and warmth.

Inheritance of function provides a different way in which conditioned reinforcers can be created. For example, many people enjoy ice cream, suggesting that, for these people, ice cream can function as a reinforcer. Through inheritance of function, any stimulus that becomes associated with ice cream in an equivalence class also may become a rein-forcer, even if the association is indirect. If you know your classical conditioning well, you may recognize this process as a possible recipe for the phenomenon called higher-order classical conditioning. Yet there are reasons to think that classical conditioning cannot fully explain conditioned reinforcement. Here are two: First, classical conditioning applies only to elicited responses such as reflexes, but not everyone would agree that “being reinforced” by a given stimulus involves a reflex. Second, higher-order classical conditioning often is weak, while conditioned reinforcement is a powerful part of the everyday world. Thinking of conditioned reinforcement as inheritance of function avoids both of these problems. Psychological effects other than classical conditioning are known to spread through equivalence classes, and the resulting effects can be quite potent.

Theories About Stimulus Equivalence

The conditions under which equivalence classes arise are becoming well understood, but theories of what equivalence is and why it occurs are still in development. Three of these emerging theories are now described briefly.

Murray Sidman (2000) proposed that stimulus equivalence is simply a complete explanation of operant learning. According to Sidman, any time operant learning occurs, all of its components parts—situational cues, responses, and the consequences that affect behavior—become members of an equivalence class. This straightforward view anticipates much of what we know about equivalence class formation in people, yet it is hard to reconcile with two types of findings. First, some studies suggest that equivalence classes sometimes form after experience that does not appear to involve operant learning, raising the question of whether stimulus equivalence is strictly an operant phenomenon. Second, it remains to be understood why nonhumans, who otherwise show operant learning quite readily, have difficulty forming equivalence classes (Zentall et al., 2002).

Two alternative views are based on the idea that equivalence-class formation is not synonymous with operant learning, but rather arises from certain kinds of operant learning histories. Naming theory states that stimulus equivalence becomes possible once learning (including operant learning) creates both receptive and expressive language skills (Horne & Lowe, 1996). According to naming theory, equivalence classes form when an individual assigns the same linguistic label (expressive repertoire) to stimuli that are not otherwise alike (the expressive language repertoire). Once assigned, this common label then prompts the individual to understand the stimuli as interchangeable in other ways (receptive language repertoire). Naming theory has some empirical support. Stimulus equivalence seems to begin in young children about the time that they become language capable, and people who initially fail to form an equivalence class may succeed after being taught a common name for the stimuli. One criticism is that naming theory is not really testable. For example, sometimes people appear to form equivalence classes without using common names for the stimuli. This result could mean that naming was absent (in which case naming theory is wrong). It could also mean that naming occurred but simply was not expressed aloud (in which case there is no objective way to verify that naming occurred).

Relational frame theory suggests that stimulus equivalence results from a skill that must be learned directly at first, and later spontaneously transfers to new circumstances (Hayes, D. Barnes-Holmes, & Roche, 2001). According to this view, early in life we learn which stimuli are interchangeable—one class of stimuli at a time, and only through the brute force of direct experience. There are no emergent relations. Thus, very young children experience the world much like the Complete Instruction Group in Zinn’s (2002) study on teaching drug names. After many stimulus classes have been mastered in this inefficient way, a generalized ability emerges. We develop a sort of conceptual template for recognizing any set of interchangeable stimuli. Now equivalence classes are readily formed, complete with emergent relations, any time that environmental circumstances expose us to overlapping stimulus relations. Note that relational frame theory is broader than the other two theories. It assumes that equivalence is just one of many possible conceptual templates for relations among stimuli (others include “opposite from,” “more than,” and “part of”). A major criticism of relational frame theory, however, is that it is vague about the type and amount of prior experience that leads to the generalized ability described.

All of these theories have stimulated important new research, but there is no agreement about which of them is most accurate. What unites these theories is an attempt to relate stimulus equivalence to simpler learning processes. As suggested by our earlier discussion of network theories of knowledge, of course, it also may be possible to think about stimulus equivalence in terms of cognitive principles of memory and concept learning.

Summary

Stimulus equivalence is one way in which individuals can “go beyond the information given” to understand things that have never been taught directly by experience. That this ability is important has never been controversial. That it arises predictably from certain kinds of experiences is only now being well understood. Research on stimulus equivalence creates the potential to engineer “reflective thinking” that is beneficial, and to understand situations in which “reflective thinking” can be counterproductive.

Much remains to be learned about stimulus equivalence, and contemporary research is unfolding on three major frontiers. First, if stimulus equivalence offers general insights into human understanding, then more studies are needed to explore its relevance to a wide range of everyday circumstances. For instance, where instruction is concerned, stimulus equivalence procedures should be useful in teaching about inferential statistics, biology, physics—any subject in which equivalences exist among terms and concepts. Also, it is important to move from merely describing pressing problems like terrorism in stimulus equivalence terms to using insights based on stimulus equivalence to address these problems in socially significant ways.

Second, the connections between stimulus equivalence and more traditional topics in psychology, such as concept learning, neural networks, and the structure of long-term memory, remain to be mapped out fully. So far, one can point to general similarities, but much research is needed to see which specific effects seen in stimulus equivalence investigations also appear in these other phenomena, and vice versa. Such research will provide a more coherent view of the intricacies of complex thinking.

Finally, and in some ways most intriguingly, research is exploring the extent to which stimulus equivalence is a uniquely human phenomenon. Where learning is concerned, nonhuman animals are surprisingly similar to us, exhibiting many of the same characteristics of operant conditioning, concept learning, and memory (e.g., Zentall et al., 2002). They do not, however, readily show stimulus equivalence. Students of psychology often wonder what it is that makes human beings special. Research on stimulus equivalence may help to answer this question.

References:

1. Barnes, D., & Roche, B. (1997). Relational frame theory and the experimental analysis of human sexual arousal: Some interpretative implications. In K. Dillenberger, M. O’Reilly, & M. Keenan (Eds.), Advances in behaviour analysis (pp. 184-204). Dublin, UK: University College Press.
2. Barnes-Holmes, D., Keane, J., Barnes-Holmes, Y., & Smeets, P. M. (2000). A derived transfer of emotive functions as a means of establishing differential preferences for soft drinks. The Psychological Record, 50, 493-511.
3. Barsalou, L.W. (1992). Cognitive psychology: An overview for cognitive scientists. Hillsdale, NJ: Erlbaum.
4. Bransford, J. D., Brown, A. L., & Cocking, R. R. (Eds.). (2000). How people learn: Brain, mind, experience, and school. Washington, DC: National Academy Press.
5. Dewey, J. (1933). How we think: A restatement of the relation of reflective thinking to the educative process. Boston: D.C. Heath.
6. Dinsmoor, J. A. (1995). Stimulus control: I. The Behavior Analyst, 18, 51-68.
7. Dixon, M. R., Dymond, S., Rehfeldt, R. A., Roche, B, & Zlomke, K. R. (2003). Terrorism and relational frame theory. Behavior and Social Issues, 12, 129-147.
8. Dougher, M. J. (1998). Stimulus equivalence and the untrained acquisition of stimulus functions. Behavior Therapy, 29, 577-591.
9. Fields, L., & Reeve, K. F. (2000). Synthesizing equivalence classes and natural categories from perceptual and relational classes. In J. C. Leslie & D. Blackman (Eds.), Experimental and applied analysis of human behavior (pp. 59-83). Reno, NV: Context Press.
10. Green, G., & Saunders, R. R. (1998). Stimulus equivalence. In K. A. Lattal & M. Perone (Eds.), Handbook of research methods in human operant behavior (pp. 229-262). New York: Plenum Press.
11. Hayes, S. C., Barnes-Holmes, D., & Roche, B. (Eds.) (2001). Relational frame theory: A post-Skinnerian account of human language and cognition. New York: Kluwer Academic/Plenum.
12. Horne, P. J., & Lowe, C. F. (1996). On the origins of naming and other symbolic behavior. Journal of the Experimental Analysis of Behavior, 65, 185-241.
13. Keenan, M., McGlinchey, A., Fairhurst, C., & Dillenburger, K.(2000). Accuracy of disclosure and contextual control in child abuse: Developing procedures within the stimulus equivalence paradigm. Behavior and Social Issues, 10, 1-17.
14. McLellan, J. A., & Dewey, J. (1908). The psychology of number and its applications to methods of teaching arithmetic. New York: D. Appleton & Co.
15. Ninness, C., Rumph, R., McCullen, G., Harrison, C., Ford, A. M., & Ninness, S. K. (2005). Functional analytic approach to computer-interactive mathematics. Journal of Applied Behavior Analysis, 38, 1-22.
16. Schusterman, R. J., Kastak, C. R., & Kastak, D. (2003). Equivalence classification as an approach to social knowledge: From sea lions to simians. In F. B. M. de Waal & P. L. Tyack (Eds.), Animal social complexity: Intelligence, culture, and individualized societies (pp. 179-206). Cambridge, MA: Harvard University Press.
17. Sidman, M. (1994). Equivalence relations and behavior: Aresearch story. Boston: Authors Cooperative.
18. Sidman, M. (2000). Equivalence relations and the reinforcement contingency. Journal of the Experimental Analysis of Behavior, 74, 127-146.
19. Stromer, R., Mackay, H. A., & Stoddard, L. T. (1992). Classroom applications of stimulus equivalence technology. Journal of Behavioral Education, 2, 225-256.
20. Zentall, T. R., Galizio, M., & Critchfield, T. S. (2002). Categorization, concept learning, and behavior analysis: An introduction. Journal of the Experimental Analysis of Behavior, 78, 237-248.
21. Zinn, T. E. (2002). Using stimulus equivalence to teach drug names: A component analysis and drug name classification procedure. (Doctoral dissertation, Auburn University, 2002). Dissertation Abstracts International, 63, 3051.

See also:

• Psychology Research Paper Topics
•  Psychology Research Paper Examples

Free research papers are not written to satisfy your specific instructions. You can use our professional writing services to order a custom research paper on any topic and get your high quality paper at affordable price.