Research Methods in Learning

Jay Brown

Research Methods in Learning

Module 02 Reading

Chapter 2

Research Methods in Learning

CHAPTER OUTLINE

Research Equipment

Classical Conditioning

Pavlov’s apparatus

Operant Conditioning

Thorndike’s puzzle box

Mazes

Radial maze

Skinner’s operant conditioning chamber

Lashley jumping stand

Motor-skills learning

Verbal Learning

Nonsense syllables

Memory drum

Review Summary

Learning Research and Experimental Design

Two-Group Design

Multiple-Group Design

Factorial Design

Review Summary

Learning in the Real World:

The study of learning is one of the most highly research-driven subject areas in psychology. For this reason, we are devoting an entire chapter to research methods that learning psychologists use (and have used) to gather the data from which they draw their conclusions. We are going to look at both the equipment and the research methodology that psychologists have used in this research enterprise.

Research Equipment

This section of the chapter will be somewhat historically oriented, as we will take a look at equipment that psychologists have used to study learning from the early years onward. To discuss the equipment that psychologists have used, we will list the type of research for which they used the equipment, but not go into detail about the research area because those areas form the basis for the rest of the text.

Classical Conditioning

As you probably remember from your introductory psychology course, Ivan Pavlov (see Figure 2-1) is credited with the discovery of classical conditioning (see Chapters 4 and 5). Pavlov was a Nobel-winning Russian physiologist who won a Nobel Prize in 1904 for his work on the digestive process. Pavlov worked with dogs in his lab, measuring the salivation reflex with test tubes attached to the dogs’ cheeks to collect saliva through a fistula. In one of the best examples of serendipity, Pavlov noticed that the dogs would often salivate before he had even presented them with some food. He termed this salivation “psychic secretions” and investigated this phenomenon for the remainder of his career, leading to his work on classical conditioning.

Interestingly, the equipment Pavlov used to study classical conditioning was exceedingly simple, perhaps even deceptively simple. As Goodwin (1991) pointed out, Pavlov’s apparatus has often been portrayed incorrectly. Figure 2-2 shows an image similar to the one that has been included in introductory psychology texts for many years. However, Figure 2-3 shows a more accurate image, based on Goodwin’s research. As you can see, Pavlov’s original apparatus or equipment barely even qualifies for that term. This state of affairs is not unusual—it is often the case that the first investigations into a topic are fairly rudimentary compared to later studies. As research areas evolve, so too does their equipment. Remember that Pavlov was originally studying the role of salivation in digestion; as he transitioned to studying classical conditioning, all he needed was to be able to collect saliva from dogs.

Operant Conditioning

You probably remember operant conditioning from your introductory psychology course—it is the process by which organisms learn to operate on some stimulus in their environment in order to achieve some purpose, often to obtain a reinforcement. We will discuss operant conditioning in detail in Chapters 6 and 7.

Thorndike’s puzzle box. Edward Thorndike (see Figure 2-4) developed the first equipment used to study operant conditioning in the late 1800s (Hilgard, 1987): a puzzle box (see Figure 2-5). Thorndike believed in trial-and-error learning, so he needed some equipment that would allow him to study this process—he made a puzzle box for cats (actually, a series of boxes). Each box was lockable and had a variety of mechanisms that the cat could manipulate in order to get out of the box and obtain a reinforcement of food. For example, the cat might have to pull a string or step on a pedal in order to release the locking mechanism and escape the box.

Thorndike noticed that cats put in the box for the first time would engage in a wide variety of behaviors before finally making the response that opened the box and allowed the cat to eat. Over repeated trials, Thorndike found that the cats tended to reduce the number and time of the unsuccessful behaviors, so that they tended to escape more quickly. After even more trials, the cats tended to engage in only the behavior necessary to escape the puzzle box and thus got out even faster. These results strengthened Thorndike’s claim of trial-and-error learning being an important form of learning.

Mazes. Thorndike also pioneered the use of mazes for studying learning, although his mazes were quite primitive and simple—he simply used books set on their sides. In 1899, Small actually introduced a more standard type of maze based on the famous Hampton Court maze (see Figure 2-6) from England (Warner & Warden, 1927). Researchers used this type of complex maze to study learning in animals such as rats and chickens. A few years later, Yerkes began using simpler mazes to study learning in lower animals such as crayfish and earthworms. The approach of studying simpler organisms led to the development of the T-maze, which has only one choice point (see Figure 2-7) and its close cousin, the Y-maze. Warner and Warden (1927) noted that at least 20 different types of organisms had been tested in mazes.

Radial maze. Olton and Samuelson (1976) pioneered the use of an 8-arm radial maze (see Figure 2-8) to assess spatial memory in rats. The interest in rats’ spatial memory arose because of their foraging behavior and food retrieval. Some species hunt or forage for food in their environment, and others hide or bury food to retrieve at a later time. Researchers were curious to discover how good rats were at remembering places that they had previously searched for food. Olton and Samuelson put food at the end of each of the eight arms of the maze and recorded the rats’ behaviors. They found that rats did have excellent spatial memories because they tended to visit different arms of the maze rather than entering an arm they had previously visited. Specifically, they found that rats entered an average of 7.6 arms in the first 8 trials.

Skinner’s operant conditioning chamber. Although Thorndike may have been the first researcher to develop equipment to use for studying operant conditioning, he and his equipment are by no means the most famous in the field. That distinction belongs to B. F. Skinner (see Figure 2-8) and his operant conditioning chamber—known by most people today as a Skinner box (see Figure 2-9). Skinner designed the operant chamber as a small enclosure for an animal (usually a rat or pigeon) in order to train it to learn a response for reinforcement—usually food, but some chambers offer the possibility of using water as the reinforcer. Figure 2-9 shows a Skinner box in which a rat would learn to press the bar in order to get a food pellet. Figure 2-10 shows a Skinner box for pigeons; the birds typically peck at a lighted disk on the wall in order to dispense some grain. We will discuss operant conditioning in much greater detail in Chapters 6 and 7.

Interestingly, the Skinner box is familiar enough that it has entered the American consciousness, something that is highly unusual for a piece of laboratory equipment. Part of the reason behind its familiarity is probably the fact that Skinner actually designed an air crib for his second daughter; Skinner’s wife wanted a safer environment for the baby than a typical crib in which infants sometimes suffocate (see Figure 2-11). Skinner wrote an article about the air crib for Ladies Home Journal—to increase attention to the article, the magazine changed the name of the article to “Baby in a Box” (http://www.bfskinner.org/brief_biography.html, a brief biography of Skinner written by his older daughter). Confusion resulted, and people believed that Skinner kept his daughter in one of his operant chambers. Rumors abounded that Skinner had in some way warped his daughter by rearing her in a box and that she had subsequently committed suicide. One way to judge the extent of common knowledge about the Skinner box is to note that you can find cartoons with a Skinner box motif (e.g., conduct an Internet search to see if you can find others besides our favorite one here—Figure 2-12).

Lashley jumping stand. In the 1930s, Karl Lashley developed the Lashley jumping stand, which he used to train rats on a discrimination task. In a discrimination task, the organism being trained must learn to distinguish between two stimuli—the stimuli can be visual, auditory, tactile, or for any sensory modality. If the organism can learn the discrimination, then the experimenter knows that the organism has the sensory capability to judge the difference between the stimuli. You have experienced many such discrimination learning situations in your life: When you learned to drive, you had to learn to discriminate many traffic signs on the basis of their shape, color, and message. You have learned to discriminate between and among hundreds or thousands of people so that you do not confuse Bob with Pete or Mary with Alice. Some discriminations are particularly difficult—if you have known identical twins only casually, you may not have been able to discriminate between them. However, as you got to know them better, you may have learned to discriminate one from the other. Discrimination tasks that are very difficult or impossible for an organism to learn let researchers know important information about the organism’s sensory abilities. For example, remember getting your eyes tested and having the optometrist ask you which stimulus is clearer? Have you ever seen day-glo orange camouflage and thought it was funny? Research has shown that deer are essentially red-green color blind (like some humans) and thus cannot see the brightly colored camouflage clothing (Murphy et al., 1993, 2001).

In research using the Lashley jumping stand, an experimenter places a rat on a stand (see Figure 2-13); the rat has a choice between two “doors” with different visual stimuli on them. When the rat jumps to the door with the correct stimulus on it, the door opens so that the rat enters a compartment with a reinforcement—usually food or water. If the rat jumps at the incorrect stimulus, the door is locked and the rat falls to the netting below. It may be necessary to encourage the rat to jump toward the doors—food deprivation is often enough motivation for a rat to explore its environment for food, so jumping off a small stand is compatible with looking for food. If the rat is hesitant to jump, the researcher may use a smaller and smaller platform or a puff of air to encourage jumping.

Motor-skills learning. Motor-skills learning refers to muscular movements “in which the amount, direction, and duration of responding corresponds to variations in the regulating stimuli” (Adams, 1987, p. 42). If you think of someone playing ping-pong, tennis, or racquetball, hitting the ball requires a myriad of adjustments in the type of shot played, most of which cannot be planned in advance, but that must take place in small fractions of a second based on the shot played by the opposing player (the regulating stimulus). Many behaviors that organisms exhibit involve aspects of motor-skills learning—eating, locomotion, fighting, sexual behavior, and so on. Even humans, as higher organisms, engage in some complex behaviors based on motor-skills learning: writing, driving, typing, assembling objects, and the like. In order to study motor-skills learning in the laboratory, psychologists have devised a couple of interesting pieces of equipment.

A mirror-tracing apparatus (see Figure 2-14) requires participants to make certain specific movements while observing their hand movements in a mirror. Thus, all movements appear reversed, which requires participants to deal with reversed feedback. The typical task is to trace a six-sided star pattern that is made up of two parallel lines such that the participant must trace between the two lines. Typical results show that, with repeated practice, people get better at the task—making few errors and taking less time. Mirror tracing mimics some real-life behaviors—notably dentistry and applying makeup.

A pursuit rotor (see Figure 2-15) requires participants to make a repetitive circular motion in order to keep a stylus held in the hand in contact with a small metal circle on a large rotating disk (somewhat like a phonograph turntable). Just as with mirror tracing, participants tend to improve (more time in contact with the metal circle) with practice. Although there does not seem to be an exact analogy in real-life to the pursuit rotor task, remember that psychological researchers are often looking for general laws about learning rather than results specific to a particular task.

Verbal Learning

Verbal learning refers to a category of learning in which researchers use verbal stimuli in order to approximate the human process of language learning. This topic is a difficult one to study because participants in verbal learning research have already learned language at an earlier age.

Nonsense syllables. Technically, nonsense syllables are not equipment in the traditional sense of the term. However, they played such a major role in the beginning of verbal learning that we are covering them. Herman Ebbinghaus, a German psychologist, pioneered the study of verbal learning and developed nonsense syllables as part of his methodology (Hilgard, 1987). As mentioned previously, studying verbal learning presented difficulties because researchers could not begin with naïve participants who were unfamiliar with the task, as researchers could with rats and mazes or Skinner boxes. To get around this problem, Ebbinghaus decided to use meaningless patterns of letters that resembled words: thus, nonsense syllables were born. Ebbinghaus used a pattern of three letters arranged in a consonant-vowel-consonant (CVC) pattern, which matches a fairly common word pattern. However, because the CVCs were devoid of meaning or previous associations, participants essentially had to learn them as if they were new words. In an interesting aside, although we used the word “participants,” Ebbinghaus used himself as his sole research participant in his various studies of verbal learning. In a sense, Ebbinghaus was also a pioneer in cognitive psychology because he was the first to demonstrate that it was possible to study higher mental processes in an empirical and experimentally rigorous manner, countering the prevailing notion that such study was not possible (Hilgard, 1987).

Memory drum. Ebbinghaus used the method of serial list learning in which he studied one CVC after another in order. Beginning his research in the 1870s (Hilgard, 1987), Ebbinghaus wrote the CVCs on slips of paper and arranged them in front of him in a list (Haupt, 2001). This method was clearly effective for learning, but it was somewhat lacking in experimental rigor. Ebbinghaus could see more than one syllable at a time and for varying lengths of time, although he did attempt to control the amount of time he viewed each syllable. These problems led Müller and Schumann to develop the memory drum (see Figure 2-16) in 1887 (Haupt, 2001). A memory drum is a machine with a rotating paper roll on which an experimenter can write or type verbal stimuli. As the roll rotates within the drum, each stimulus is exposed in the “window” of the drum for a fixed amount of time—only one stimulus is visible at a time, and all stimuli have identical presentation times. Thus, the memory drum allowed greater experimental rigor in verbal learning research than Ebbinghaus used.

Review Summary

Pavlov began the study of classical conditioning, but the equipment was extremely simple. Many books have incorrectly depicted his apparatus.

The study of operant conditioning has entailed the use of Thorndike’s puzzle boxes for cats, various types of mazes, Skinner’s operant conditioning chamber, Lashley’s jumping stand, and motor-skills apparatuses such as mirror-tracing and pursuit rotor equipment.

Ebbinghaus began the study of verbal learning to mimic human language learning and used nonsense syllables to allow learning of new verbal stimuli. Müller and Schumann developed the memory drum to present verbal stimuli one at a time in serial order.

Learning Research and Experimental Design

As you can probably tell by content of this chapter to this point, the field of learning involves a great deal of experimental research. We believe that understanding the research process is vital to more fully understanding the field of learning. To that end, we are including in this chapter information about experimental design and how it can interface with learning research.

Figure 2-17 features a research design flowchart that you can use to help conceptualize experiments that you read about this semester. If you begin at the left side of the flowchart and answer a few simple questions, you will be able to find out what type of research design a particular experiment used. As you will see, different types of research designs allow researchers to draw different types of conclusions. To help make the subject of research design more concrete for you, we will use a hypothetical learning experiment to trace throughout the flowchart.

Two-Group Design

Suppose Nancy, a student researcher, wants to find out how caffeine affects learning in rats. (This hypothetical example is based on actual research by Cathey, Smith, and Davis, 1993.) In experimental terminology, caffeine would be the independent variable, and learning would be the dependent variable.

***** Do you know why caffeine is the independent variable and learning the dependent variable rather than vice versa?

—————————–

The independent variable is the variable controlled and manipulated by the researcher—in this case, Nancy will control the caffeine exposure for the rats. The dependent variable is the variable measured by the experimenter to determine the effects of the independent variable. An extraneous variable is any variable that might affect learning other than caffeine—it is the experimenter’s task to eliminate or remove the possible effects of extraneous variables from the experiment. It is often the case that the experimenter will have to develop operational definitions for the variables. In this way, the experimenter not only develops precise definitions of what “caffeine” and “learning” mean, but also develops specific explanations of how each variable is manipulated or measured. Nancy is interested in caffeine’s effects on maturing rats, so she specifies that the operational definition of caffeine for her experiment is .17 mg/ml (.017%) of caffeine in drinking water throughout gestation, weaning, and early adulthood (the time of testing). Note that this operational definition is quite precise—another researcher could replicate it without problem. She decides that she will have one group of rats that receives caffeine (experimental group) and one group that does not receive caffeine (control group). Nancy has many different choices as she thinks about her operational definition of learning. She decides to time how long it takes the rats to learn to bar press. Further, she operationally defines “learn to bar press” as a rat making 10 bar presses within a 30-second period. Again, this definition is precise enough that Nancy can easily measure learning and that another researcher could replicate it. Finally, Nancy has to decide where she will get her rats. Because she is interested in caffeine’s effects before and after gestation, she will need to expose the rats to caffeine before they are born. She gets four female rats and randomly assigns two to the experimental group and two to the control group. She supplies them with the appropriate type of water and then breeds the females. When the rat pups are born, she will have two litters in the experimental group and two litters in the control group.

Figure 2-17 provides you with the information you need to classify Nancy’s experimental design.

***** Can you figure out Nancy’s research design and what statistical test she would use to analyze her data?

—————————

Ideally, you followed these steps and came to the following conclusions. Nancy has only one independent variable (IV) in her study: the caffeine. Although she has two groups (caffeine and no caffeine), these are simply two levels of her IV, not two different IVs. Thus, Nancy’s experiment will have 1 IV with 2 groups. The final question you had to answer deals with the two subject groups—whether they are independent or related. Because Nancy randomly assigned the mother rats to the two conditions, the groups of pups are essentially randomly assigned also. The two groups are not related because Nancy did not use any of the three methods of creating related groups: The rat pups do not participate in both the caffeine and no-caffeine conditions, so there are no repeated measures; Nancy did not attempt to form matched pairs of rat pups in the two conditions, so there is no matching; and, although she will use rats from litters, all the pups from a litter will be in the same group, so natural pairs is not possible either. Therefore, following the flowchart question by question, Nancy will use a 2 independent groups design and will analyze her data with a t test for independent samples. (If Nancy used related groups, she would analyze her data with a t test for correlated samples.) If you have taken a statistics or research methods course, these terms should be familiar to you. If not, you will learn more about them when you take those courses.

Suppose Nancy runs her experiment, analyzes the data, and finds that the data show no effect of caffeine on how quickly the rats learn to bar press. She is surprised because her background research showed that caffeine tends to increase arousal and task performance (e.g., Glavin & Krueger, 1985). However, in reviewing this research, she sees that several researchers have used varying concentrations of caffeine in their studies, which makes her wonder if the amount of caffeine she used was adequate to have any effect on learning.

Multiple-Group Design

Because Nancy is uncertain about whether the caffeine adequate, she decides to conduct another experiment using varying amounts of caffeine. Because she found no difference in her two groups in the first experiment, she sees no need for another control group, but she does want to contrast more than two different amounts of caffeine. She decides to use the caffeine concentration from her first study as the minimum amount in the new study and to compare the effects of that concentration to amounts that are two and three times as much.

***** Can Nancy use the same experimental design and statistical test as she did for the first study?

————————–

Again, Figure 2-17 can help you answer this question. Does Nancy still have only one IV? Yes—her IV is still caffeine administration. Does Nancy still have two groups or levels of her IV? No—she has now chosen to have three groups: 0.017, 0.034, and 0.05% caffeine groups. The flowchart shows that a different path is necessary with three (or more) groups. Nancy must also answer the question about what kind of participant groups she will use. She decides that she will obtain her rats for this study in much the same manner that she did previously—so, again, the three groups will be independent. Therefore, for her follow-up experiment, Nancy will use the multiple independent groups design and analyze her data with a one-way ANOVA for independent groups.

Suppose that Nancy now conducts her experiment, collects and analyzes the data, and finds significance in her findings from the one-way ANOVA. She consults her statistics book and refreshes her memory that significance in a one-way ANOVA means that there is a significant difference somewhere among her means. To determine which means are significant from each other, she must conduct post hoc tests, such as Tukey’s HSD. After she conducts her post hoc tests, Nancy finds out that the .05% group shows a significant difference in learning from the other two groups—the .05% group has learned to bar press more slowly than the other two caffeine groups.

*****Suppose that Nancy had found a significant difference in her first experiment. Would post hoc tests have been necessary to tell how the groups differed?

—————————–

In a two-group experiment, post hoc tests are not necessary. With only two groups in the experiment, a significant difference can only mean that one group has scored lower or higher than the other group. With three (or more) groups, however, a significant ANOVA can mean that any group (or groups) is different from any other group (or groups).

Factorial Design

Nancy now has some evidence that a higher dose of caffeine may inhibit learning in rats. However, having been a good student in her statistics and research classes, she knows that she should not jump to a conclusion too quickly. In looking at the research literature, she found that Holloway and Thor (1982) found that caffeine had differing effects on locomotor activity over time in rats. They found that caffeine produced an early increase in movement that was followed, over time, by decreased movement. Nancy begins to wonder if the bar pressing results from her second experiment might have been as much a product of movement as of the caffeine. For this reason, she plans yet another experiment.

Because of her uncertainty about the caffeine effects, Nancy decides to go back to using only a control and experimental group, but she will use the higher dose of caffeine from Experiment 2 that did seem to cause an effect. In addition, because of the Holloway and Thor (1982) research, Nancy decides to add a second IV—learning measured over time. If caffeine has activity effects that change over time, then measuring over repeated trials is a good way to get a handle on whether the effects on learning would also change over time. Finally, Nancy also decides to add a second measure of learning; in addition to bar pressing, she decides to also use a maze because of running a maze is more consistent with the locomotor activity that Holloway and Thor (1982) measured.

*****Based on this description of Nancy’s new experiment, what design (and statistical analysis) will she use?

——————————-

Of course, you should turn to Figure 2-17 to answer this question. If you read the last paragraph carefully, there was a big clue given—did you find it? We said that Nancy was adding a second IV, so you can see that the answer to the first question takes you in a different direction on the flowchart than Nancy’s previous research questions. Anytime a researcher uses more than one IV, the research design becomes a factorial design, which simply means that the design has more than one factor (IV). Also, notice from Figure 2-17 that the question about the number of levels or groups of the IV is not relevant when you have two or more IVs—the design will always be a factorial design. Thus, the only other question you have to answer after specifying two IVs is the kind of participant groups. In her other two experiments, Nancy has used independent groups for the caffeine IV, so it is a logical assumption that caffeine will again be a between subjects factor (two randomly assigned groups). The second IV Nancy is using is learning over time—in other words, she will measure the groups of rats repeatedly across several trials. This type of IV creates related groups—in this case, repeated measures, because each rat is measured multiple times. So, Nancy’s experiment will have two IVs—one with independent groups (caffeine) and one with related groups (repeated trials). Therefore, Nancy’s design is a factorial mixed-groups design; she will analyze her data with a factorial ANOVA for mixed groups.

Notice that Figure 2-17 makes no reference to the number of dependent variables (DVs) in choosing an experimental design. Although Nancy will measure the results of the caffeine administration on both bar pressing and maze learning, she is not attempting to analyze both DVs at the same time. Thus, she will analyze the bar press data with one factorial ANOVA for mixed groups and the maze data with a second identical analysis.

In real life, this is exactly the experiment that Nancy Cathey, an undergraduate student at Ouachita Baptist University (Arkadelphia, AR), conducted under the supervision of her faculty sponsor Randolph Smith and consulting faculty sponsor Stephen Davis. In the experiment, Cathey et al. (1993) found that rats made more bar press responses over eight trials of 8 minute periods. In other words, the rats learned to bar press better over time—not a surprising result. However, the rats that ingested caffeine made fewer bar presses than rats that did not ingest caffeine beginning on Trial 3 through the end of the experiment. Thus, for bar pressing, caffeine inhibited learning.

In the maze task, rats that ingested caffeine completed a higher percentage of the maze in 8 minutes on the first two trials compared to rats that had not ingested caffeine. However, over trials, that advantage disappeared by Trial 3; on the last four trials, noncaffeine rats completed 100% of the maze whereas caffeine rats completed slightly less than 100% (although the differences were not significant). Cathey et al. (1993) also timed the rats in the maze, so it was possible to compare maze completion time between the groups. The results showed that caffeine rats took less time to complete the maze on Trials 1-4, although the difference was significant only on Trial 4. However, the results then reversed, with the noncaffeine rats taking less time to complete the maze on Trials 5-8, with those differences showing significance on Trials 5, 6, and 7.

Experimental Design Summary

We hope that this brief look at how learning researchers use different experimental designs to ask different questions will help you better understand the research studies that we cite in the remainder of the book. There is nothing magical or mystical about learning research—it follows standard rules of research that apply to all areas of psychology. We believe, however, that it is difficult to understand the research process if you do not understand the basic building blocks of research—experimental design.

Review Summary

Learning researchers conduct experiments to determine how various factors affect the learning process. Experimental design refers to the various considerations that researchers must take into account as they plan an experiment.

A researcher manipulates an independent variable to determine its effects on a dependent variable. The researcher must control or eliminate extraneous variables so that they do not influence the dependent variable.

Two-group designs consist of one independent variable that has two amounts or types. Often in a two-group design, one group of participants (experimental group) will receive the independent variable and one will not (control group).

A researcher can assign participants to groups such that the groups are independent or related. Independent groups come from random assignment to groups whereas related groups come from one of three techniques: repeated measures, matched groups, or natural groups.

A two-group design with independent groups requires a t test for independent groups to analyze the data. A two-group design with related groups requires a t test for correlated groups.

Multiple-group designs consist of one independent variable that has more than two amounts or types. A multiple-group design may or may not have a control group. Groups in a multiple-group design may be either independent or related.

A multiple-group design with independent groups requires a one-way ANOVA for independent groups. The same design with related groups requires a one-way ANOVA for correlated groups.

Factorial designs consist of two independent variables, regardless of the number of levels of the independent variables. Groups in a factorial design may be independent (for all IVs), related (for all IVs), or a mixture of the two.

A factorial design with all independent groups requires a factorial ANOVA for independent groups, with all related groups requires a factorial ANOVA for correlated groups, and with a mixture of groups requires a factorial ANOVA for mixed groups.

Learning in the Real World: Transfer from the Laboratory

Sometimes students in a learning course get fixated on the course material and “don’t see the forest for the trees.” In this context, we are referring to students who get hung up on the fact that much learning research has taken place with animal subjects and then complain about their perceived lack of relevance of the material to their lives. These students seem to think that the course topic is animal learning rather than learning in general. We can assure you that if you take this perspective, the course will likely be more difficult than if you take the broader view.

As you go through this course, you need to keep in mind yet another important experimental concept—that of generalization. In most cases, researchers do not conduct experiments to study just the participants in the actual experiment. Instead, the researchers hope that their results will apply to (generalize to) other organisms and other situations or environments. If the results do not generalize beyond the experiment, then the research does not show external validity. If you keep generalization in mind when you read about learning research, you should realize that, although a research study may focus on learning in rats or dogs or pigeons, the results may well generalize to humans. In fact, to help you avoid some students’ problem with a fixation on animal learning, it should be your challenge to try to come up with a generalization for each result of animal research to a similar situation for humans. In this “Learning in the Real World” section, we will “prime the pump” by providing you with some examples that correspond to the Research Equipment section of this chapter.

As you probably remember from earlier in the chapter, classical conditioning has been studied without much in the way of equipment, so you should not be surprised to find examples of classical conditioning in many aspects of your life. If you have experienced some important stimulus in your life paired repeatedly with a neutral stimulus, then classical conditioning has probably occurred. Have you ever wondered why an old song makes you think of an old boyfriend or girlfriend? Perhaps that was “your song” together, so the song continues to make you think of that person despite the fact that you have long since split up. Have you ever noticed that suspenseful or spooky or scary music in a movie or TV show leads you to expect something dramatic is about to happen? This expectation developed because producers use music to set the stage for important events to occur. You have become classically conditioned to expect something to happen after you hear a certain kind of music. Wouldn’t it be nice if the same thing happened to you in real life so you would have a warning?!

We covered a variety of different types of operant conditioning paradigms and equipment in this chapter. It is interesting to attempt to find real-life examples of those situations. Remember Thorndike’s puzzle boxes? When I was growing up, my grandfather had a small puzzle box in which you could hide something—it required a series of manipulations to open, just like the cats had to learn. If you have never seen such boxes, simply type “puzzle box” into your favorite Internet search engine, and you will find an amazing array of such boxes. You are probably familiar with mazes from your childhood—many children’s magazines and activity books have mazes of varying difficulty for children to trace. As an adult, you have encountered learning similar to maze learning when you tried to figure out the best route to get from your home to school or other destinations. Now, when you drive from Point A to B, the process is so automatic that you probably don’t even realize it. However, if there is road construction or you have to detour for some other reason, the process of getting through the “maze” becomes much more conscious.

Obviously, you have never found yourself in a Skinner box or on a Lashley jumping stand, but you have used the principles from those situations in many different learning situations. In reality, you have behaved very much like a rat pressing a bar when you have used a vending machine to obtain a soda or a snack—the similarity is striking when you think about it. The main principle behind the Skinner box is operant conditioning itself—where you make a certain response in order to achieve a particular outcome. You inserted money and then pressed a button in order to get your reinforcer. Lest you think operant conditioning applies only to situations that closely resemble a Skinner box, you should realize that you have relied on operant conditioning to learn a wide variety of behaviors, from writing and typing to driving a car to combing your hair to tying your shoes. The general principle from the Lashley jumping stand is discrimination learning—gaining the ability to tell the difference between two or more stimuli. Discrimination is critical to your survival—depending on your environment, you may have to learn the difference between different traffic signals or signs (the difference between red and green lights or octagonal [stop] and triangular [yield] signs can be life threatening!), a venomous coral snake and a safe milk snake (“red and yellow, kill a fellow; red and black, friend of Jack” is a common mnemonic device used to tell whether a snake is venomous or not based on touching color bands), or edible and poisonous mushrooms.

The final two types of learning scenarios we covered earlier are motor-skills learning and verbal learning. We gave examples of these types of learning at that point in the text, so will not go further at this point. One point of clarification about motor-skills learning is in order, however. You may have noticed that we listed some of the same learned behaviors (writing, typing, driving) under both operant conditioning and motor-skills learning. This double listing was not an error—it is important to realize that some types of learning may double up. If we take driving as an example, there is a lot about driving that involves motor skills: starting a car, using turn signals, coordinating your feet to use the gas and brake pedals, and so on. If you have ever driven (or tried to drive) a car with standard transmission, you realize just how important motor-skills learning is; using a clutch is a tricky proposition! Another good example of motor skill involved in driving is using the steering wheel—just watch kids as they pretend to drive—turning the wheel wildly as they typically do would result in many wrecks. At the same time, a great deal of the knowledge you have about driving is not motor skills. For example, you probably had to read a driver’s manual before you began to drive so that you would know various laws, traffic signals, and other important information. Likewise, the person who taught you to drive probably talked a great deal about driving before ever letting you behind the wheel.

We hope these examples have helped you see how learning research that originated primarily with animal subjects also applies to humans. Again, we urge you to work on your powers of generalization as you read through the remainder of this text.

Margin Definitions

classical conditioning

(undoubtedly defined in another chapter—copy from there) a form of learning in which an organism learns to associate a stimulus with a response

serendipity

an accidental discovery—a scientist studying one concept happens to come across another concept, which is actually more important than the original concept

operant conditioning

(undoubtedly defined in another chapter—copy from there) a form of learning in which an organism learns to make a response in order to obtain a reinforcement (or avoid a punisher)

reinforcement

(undoubtedly defined in another chapter—copy from there) something that follows a behavior that strengthens that behavior and increases its future probability

trial-and-error learning

learning in which organisms make a number of incorrect responses before they are successful and happen to make the correct response

puzzle box

an experimental apparatus pioneered by Thorndike; cats had to learn how to manipulate stimuli in the box to be released from the box

maze

an experimental apparatus pioneered by Thorndike; usually involves a trail that an organism must learn to navigate without getting lost

T-maze

a simple maze shaped like a T; organisms can only turn right or left at one choice point

Y-maze

a simple maze shaped like a Y; organisms can only veer right or left at one choice point

8-arm radial maze

a maze that resembles an asterisk (*) with 8 arms used to test spatial learning; the organism starts in the middle space and obtains food reinforcement at the end of each arm

spatial memory

memory that involves items or directions in space—the type of memory used to learn mazes

operant conditioning chamber

a small box in which an animal learns a task in order to obtain a reinforcement—often, a rat pressing a bar to get a food pellet; also known as a Skinner box

discrimination task

a learning task in which an organism must learn to distinguish between two (or more) stimuli

motor-skills learning

a type of learning that involves bodily movements (e.g., writing, playing tennis, throwing)

mirror-tracing apparatus

an experimental apparatus that requires a person to trace a figure (usually a star) while viewing that figure in a mirror

pursuit rotor

an experimental apparatus that requires a person to use a stylus to maintain contact with a small point on a rotating disk

verbal learning

a type of learning pioneered by Ebbinghaus that entails a participant learning verbal materials (e.g., words, sentences)

nonsense syllables

a type of verbal material that is devoid of meaning used in verbal learning; pioneered by Ebbinghaus

CVC

a common type of nonsense syllable in a consonant-vowel-consonant pattern

memory drum

a mechanical apparatus used to present lists of words (or other verbal material) in a sequential fashion

experimental design

a general plan for selecting participants, assigning participants to experimental conditions, controlling extraneous variables, and gathering data

independent variable (IV)

a stimulus or aspect of the environment an experimenter manipulates to determine its influence on behavior

dependent variable (DV)

a response or behavior that the experimenter measures; when the IV is significant, changes in the DV are directly related to the manipulation of the IV

extraneous variable

undesired variables that may operate to influence the DV and, thus, invalidate an experiment

operational definition

defining the independent, dependent, and extraneous variables in terms of the operations used to produce them

experimental group

in a two-group design, the group of participants that receives the IV

control group

in a two-group design, the group of participants that does not receive the IV

random assignment

a control technique that assures that each participant has an equal chance of being assigned to any group in an experiment; used to equate groups

levels

differing amounts or types of an IV used in an experiment (also known as treatment conditions)

independent groups

groups of research participants that are formed by random assignment

related groups

groups of research participants that are related in some way—through matching, repeated measures, or naturally occurring sets

repeated measures

an experimental procedure in which research participants are tested or measured more than once

matched pairs

research participants in a two-group design who are measured and equated on some variable before the experiment

natural pairs

research participants in a two-group design who are naturally related in some way (e.g., a biological or social relationship)

2 independent groups design

a research design with two groups of participants that are formed by random assignment

t test for independent samples

an inferential statistic used to evaluate the difference between two means from randomly assigned groups

t test for correlated samples

an inferential statistic used to evaluate the difference between two means from related groups

multiple independent groups design

a research design with more than two groups of participants that were formed by random assignment

one-way ANOVA for independent groups

a statistical test used to analyze data from an experimental design with one IV that has three or more groups that were formed by random assignment

one-way ANOVA for correlated groups

a statistical test used to analyze data from an experimental design with one IV that has three or more groups that are related (i.e., repeated measures, matched groups, or natural groups)

post hoc test

a statistical comparison made between three or more group means after finding a significant difference among the groups

Tukey’s HSD

a post hoc test commonly used to compare differences among three or more groups that showed a significant difference

factorial design

an experimental design with more than one IV

factorial mixed-groups design

an experimental design with more than one IV; at least one IV has groups formed by random assignment and at least one IV has related groups

factorial ANOVA for mixed groups

an inferential statistical test used to test data from a factorial mixed-groups design

generalization

applying the results from an experiment to a different situation or population

external validity

a type of evaluation of an experiment; do the experimental results apply to populations and situations that are different from those of the experiment?

mnemonic device

a device such as a saying, rhyme, or trick to help remember information; for example, the saying “i before e, except after c” to help remember a difficult spelling rule

————————————————————————————————————

References

Adams, J. A. (1987). Historical review and appraisal of research on the learning, retention, and transfer of human motor skills. Psychological Bulletin, 101, 41-74.

Cathey, N. R., Smith, R. A., & Davis, S. F. (1993). Effects of caffeine on rats’ barpress and maze performance. Bulletin of the Psychonomic Society, 31, 49-52.

Glavin, G. B., & Krueger, H. (1985). Effects of prenatal caffeine administration on offspring mortality, open-field behavior and adult gastric ulcer susceptibility. Neurobehavioral Toxicology & Teratology, 7, 29-32.

Goodwin, C. J. (1991). Misportraying Pavlov’s apparatus. American Journal of Psychology, 104, 135-141.

Haupt, E. J. (2001). The first memory drum. American Journal of Psychology, 114, 601-622.

Hilgard, E. R. (1987). Psychology in America: A historical survey. San Diego: Harcourt Brace Jovanovich.

Holloway, W. R., & Thor, D. H. (1982). Caffeine sensitivity in the neonatal rat. Neurobehavioral Toxicology & Teratology, 4, 331-333.

Murphy, B.P., Miller, K. V., Marchinton, R. L., Deegan, J., II, Neitz, J., & Jacobs, G. H. (1993). Photopigments of white-tailed deer. Southeast Deer Study Group, 16, 32.

Murphy, B.P., Miller, K. V., Marchinton, R. L., Deegan, J., II, Neitz, J., & Jacobs, G. H. (2001). Photoreceptors and daylight vision of the deer. In A. Sivic & L. E. Sielecki, Wildlife warning reflectors spectrometric evaluation. Environmental Management Section, B. C. Ministry of Transportation and Highways, Victoria, British Columbia.

Olton, D. S., & Samuelson, R. J. (1976). Remembrance of places passed: Spatial memory in rats. Journal of Experimental Psychology: Animal Behavior Processes, 2, 97-116.

Warner, L. H., & Warden, C. J. (1927). The development of a standard animal maze. Archives of Psychology, No. 93, 1-35.

Figure 2-1

Ivan Pavlov