Scientific Psychology and Its Research Methods

Introduction to Psychological Science: Integrating Behavioral, Neuroscience and Evolutionary Perspectives - William J. Ray 2021

Scientific Psychology and Its Research Methods


2.1 Summarize how science can help us be part of and explain our world.

2.2 Describe the major research designs used in psychological science.

2.3 Explain the techniques used to understand and evaluate scientific data.

2.4 List the steps involved in designing an experimental study.

2.5 Evaluate the ethical guidelines that psychologists follow when conducting scientific research.

Neuroscientist V. S. Ramachandran treated an individual named David at his medical center in San Diego, California (Ramachandran, 1998). David had no problems with memory, engaged easily in conversation, expressed emotions, and otherwise appeared “normal.” However, he did one very puzzling thing. He did not recognize his own mother. When she appeared, he insisted, “That woman looks exactly like my mother, but she is not my mother!”

As a health care professional, how might you understand this? You might ask if this was some type of psychosis in which he had the delusion that his mother was not his mother. However, David showed no other signs of disorganization or problems with functioning. You might also ask if David had any type of emotional conflict with his mother. The answer to that was also no. After more information gathering, it was discovered that David at times also thought his father was not his father. Eventually, it was revealed that David did indeed experience his parents as really his parents when talking to them on the phone. Seeing them resulted in a different experience.

The formal name for this condition is the Capgras syndrome, named after the physician who first described the symptoms in the 1920s. However, the mechanisms involved were not known. Since David had previously had a motorcycle accident, it was possible that normal brain processes were not functioning correctly. In order to understand David, Ramachandran asked himself what was missing in David’s experience of his mother. His answer was that there was no emotional response. How can we test the emotional response to seeing a face? Emotion is not only processed in the brain but also in the autonomic nervous system (ANS), which prepares the body for dangerous situations. For example, if we see a bear and start running, it is the sympathetic part of the ANS that makes us feel excited and moves blood to our muscles for a quick getaway. One easy way to measure the sympathetic nervous system is to pass a small electrical current along the skin, usually between the palm and the finger. This procedure is referred to as electrodermal activity (EDA), which has also been used in lie detection. If we are excited, then our skin sweats slightly, this in turn, makes it easier for the electrical current to pass between the two electrodes. Whenever we have an emotional response to what we see, we get changes in the EDA. Not surprisingly, David did not show any EDA differences when viewing pictures of those close to him. This suggested to Dr. Ramachandran that there was a disconnect between David’s visual face perception areas and the emotional centers of the brain. Since the auditory system connects to emotional centers of the brain differently than the visual system, then it makes sense that David did not have the same emotional disconnect when talking with his parents on the phone as when he saw them.

To study an unknown condition, as scientists we must logically examine important components of the condition. In David’s case, what was important was both what he experienced—not believing his mother was his mother—as well as what he didn’t experience—the emotional response to his mother.

What Is Science?

Science is a method for asking questions about the world. The quality of the answers we receive is influenced by several factors, one of the most important being the experimental design that we use. A well-controlled experiment allows us to be more certain of our results. As a scientist you could also ask other questions about David’s case. You might begin by investigating what the research literature tells us about the cognitive and emotional factors involved in recognizing and experiencing a parent. As a scientist, you would also want to know:

✵ What factors are related to experiencing a parent and what factors can be ruled out?

✵ How would you go about studying the phenomenon of Capgras syndrome?

✵ What techniques can you use to measure cognition and emotion?

This is a general process called science that involves experiencing the world and then drawing general conclusions, or facts from observations. Sometimes these facts are descriptive and can be represented by numbers. For example, we know that the moon is 238,000 miles from Earth and that the brain uses 20% of the energy produced by the body. Other times, knowledge is more general and can describe a relationship or a process, like the fact that it is more difficult to learn a second language after puberty than before, or that following puberty, most adolescents want to be with their friends rather than their parents. Whatever the topic, the known facts about a particular subject are called scientific knowledge (see Ray, 2022 for more information).

A Human Activity

Science is above all a human activity. Some people perform science as their profession and are known as scientists. Equally accurate is the idea that all people perform science in some form. In fact, science in many ways is similar to the way we have been learning about the world since we were infants. We learn through interacting with our world. In fact, you probably know much more about the scientific method than you think you do. You have been using it in one form or another since you first began toddling about and discovering the world.

Watch a young child. When something catches his or her eye, the child must examine, observe, and have fun with it. Next, the child wants to touch it. Some interactions are fun: “If I tip the glass, I get to see the milk form pretty pictures on the floor.” Others are not so much fun: “If I touch the red circles on the stove, my fingers hurt!” From passive observations to active interactions, the child learns a little more about the world.

Figure 2-1 Child playing with blocks and learning about simple principles of physics and spatial abilities.

Figure 2-1 Child playing with blocks and learning about simple principles of physics and spatial abilities.

Like the child, scientists are exploring the unknown—and sometimes the known—features of the world. All basic research strategies are based on one simple notion: To discover what the world is like, we must experience it. To have an idea about the nature of the world is not enough. Instead, like the child, scientists experience the world to determine whether their ideas accurately reflect reality. Direct experience is an essential tool because it alone allows us to bridge the gap between our ideas and reality.

Science and Doubt

There is another aspect to science that many people do not think about. This is the aspect of doubt. In science we use doubt to question our ideas and our research and ask whether factors other than the ones that we originally considered might have influenced our results. By doing this, we come to see that science is a combination of interaction with the world and logic. The logic of science leads us to the realization that one of the real strengths of science is showing us when we are wrong. If someone says that all swans are white, for example, seeing a white swan—or thousands of swans, all of which are white—does not actually prove this to be the case. However, a black swan would clearly show that the statement was wrong. In this spirit, Einstein is reported to have said, “No amount of experimentation can ever prove me right; a single experiment can prove me wrong.” The philosopher of science Sir Karl Popper referred to this approach as falsification. Thus, one important aspect of doing science is to ask yourself, how would I know if I was wrong? By the way, being wrong also helps researchers to know factors that are not involved.

As you think about falsification and doubt, you begin to realize that doing psychological science is a broad and critical process. It is more than designing an experiment and looking at the results. It is also thinking critically about the experiment and the results. As you will see in this chapter, at every step of performing psychological science, you ask if there could be other explanations for understanding the results. You can also ask if there are other studies that would help you support or better understand your results. You use logic, scientific methods, and your own experience to critically evaluate your ideas.

Debunking Superstition and Pseudoscience

One of the goals in this book is to help you think critically. This is done in the context of psychology as a science. Science emphasizes ways to evaluate information and come to valid conclusions. Experience and doubt form two important characteristics of psychological science.

However, other aspects of our human nature may work against this. Human beings have a long history of relying on magic and superstition to guide our decisions. Even today some hotels do not have a 13th floor, since this is considered by some to be unlucky. Some people carry lucky charms or wear certain clothes to important meetings because they think this will help them succeed. Even our newspapers have daily horoscopes to tell us what types of activities to engage in that day.

Most likely, superstition is part of our long evolutionary history. We may emotionally believe that something is true and hate the thought of giving up that particular idea or belief. Social and cognitive psychology research shows that this is made even more complicated by the fact that when we believe something to be true, we tend to look only for supporting evidence and to dismiss any contradiction to our belief as an exception (Kahneman, 2011). Without a method to test our ideas, we will never know the validity of our conclusions. That is one important aspect of science.

You also know that frequently individuals make claims in the name of science that are not actually based on rigorous scientific procedures. Often these are individuals who want to sell you something. You see this every day on social media. Some of these claims have made their way into popular books and even have been reported as true in the mass media. For example, the idea that you can learn while you sleep was a popular one during parts of the 20th century. This idea found its way into a variety of movies and novels and was even spoofed on an episode of The Simpsons.

The Simpsons had it correct, since numerous scientific research studies have not been able to find evidence supporting the claim of learning during sleep (Simon & Emmons, 1955). Despite this, even today you can go on the web and buy materials that claim to help you learn languages, study for exams, and improve a variety of abilities while you sleep. All of these items are supposed to be based on scientific research, or at least that is what the websites say.

The phenomenon of presenting information as if it is based on science when it is not is referred to as pseudoscience. Pseudoscience is false science. Often claims of pseudoscience are based on testimonials that present a single side of the issue and have not been evaluated by others studying similar phenomena using valid scientific methods. For this reason, scientists pay particular attention to research that has been evaluated by other scientists before it is published. This process is called peer review, and journals that follow this procedure are peer-reviewed journals. One characteristic of pseudoscience is that it is not found in peer-reviewed journals. As we continue throughout the book, you will learn how to identify pseudoscience and develop some of the skills of critical thinking to evaluate its claims.

Just because something is not based on science does not mean that we shouldn’t consider it. Many of us love to read science-fiction and other types of fantasy literature. Such writing makes us consider possibilities and other ways of thinking about our world. As we will see in this book, scientists are always considering alternative explanations and sometimes ideas that seem crazy. However, considering that something could be true does not mean that it has scientific support. Thus, we need a means for testing our ideas. We particularly need ways of knowing if we are wrong, which is one of the important aspects of designing scientific research.

Unintended Influence

Besides pseudoscience, sometimes even the person making the claim believes it to be true. “There is a horse that solves correctly problems in multiplication and division.” This is the provocative statement that opens a book (Clever Hans) that describes the remarkable story of a German horse nicknamed Clever Hans (Pfungst, 1911). People from all over Germany wanted to see this horse as it tapped out the correct answers to math problems. It could even solve problems that involved fractions by first tapping the numerator and then the denominator. If that was not enough, Hans also knew the yearly calendar and was able to give you the date of any day you would present.

As you also can imagine, people who knew horses well and how to train them were skeptical. The book’s introduction asserts that thousands of these individuals watched Hans and his owner, a Mr. von Osten, over many months and could not find a trick. Mr. von Osten had assumed that animals were smart and could be taught if only given enough time.

Mr. von Osten taught Hans as if he were a German schoolboy. There were blackboards, flash cards, and all the curriculum subjects available in Germany. Over a period of years, van Osten taught Hans to move his hooves in a way to differentiate between yes and no, and to tap out numbers. Food, praise, and other rewards were given for correct responses. In the Germany of the early 1900s, some thought this was a way to understand animal consciousness. Others thought it was all a fraud. Thus, we are left with a riddle, as the book says.

In order to solve the riddle, Oskar Pfungst, a German psychologist, began a series of experiments that were published in 1911. Pfungst first tells us that we cannot determine Hans’s abilities in the ordinary conditions in which he performed. This is because he could not control the environmental factors surrounding the performance. In order to reduce the environmental influences, Pfungst erected a large canvas tent within Mr. von Osten’s courtyard in which the horse could freely move. Pfungst then tried to have the horse respond to the questions of others. In fact, he tried some 40 different individuals with inconclusive results.

What does this mean? It means that there is something special about Mr. von Osten that allows the horse to answer. Oskar Pfungst then did a very clever experimental manipulation. He alternated questions to which Mr. von Osten knew the answer with those he did not. If Hans was able to answer both types of questions, then it could be concluded that the questioner was not influencing the horse. If Hans was only able to answer questions that the questioner himself knew the answer to, then Pfungst could begin to look for subtle influences on Hans. To further restrict these influences, Pfungst placed blinders on Hans so he could only see in front of him.

What do you think was found? Pfungst found that if the questioner knew the answer to the question, then Hans was correct nearly all the time. If the questioner did not know the answer, Hans was rarely correct.

How did Mr. von Osten communicate the correct answer to the horse? He did what all of us do when we ask a child to count. We lean forward slightly when the child is answering the question and then lean backwards when the correct answer is given. Hans had learned to keep tapping when the questioner was leaning forward and to stop when he leaned backwards. Mr. von Osten had no idea that he was not teaching the horse to count during the years of training but rather training him to respond to subtle cues that neither he himself nor those watching him work with the horse could see.

The story of Clever Hans shows us that all organisms, including humans, are sensitive to their environment. This is true whether we are interacting with other people or being part of an experiment. Thus, in psychological experiments there are a number of opportunities for those who are researchers to influence those who are participants in ways that were never intended. Participants can also influence researchers. For example, participants in experiments may respond to their own ideas or internal demands rather than those of the experiment itself.

What Are the Methods of Science?

The methods of science closely parallel our ways of learning about the world. We interact with the world and gain new knowledge and understanding of what we experience. However, in doing psychological research, we do this in a more organized and logical manner. We may observe and describe what we see, which is referred to qualitative research. We can also measure and report the number of specific factors that can be used in a mathematical manner, which is referred to as quantitative research. The scientific approach emphasized in this chapter will describe both qualitative and quantitative approaches. In both, researchers approach a particular question or topic in a logical manner. We can think about a critical approach to science in terms of four stages.

1. Scientists first begin with an idea or expectation. A formally stated expectation is called a hypothesis. The scientist says, “I expect this to happen under these conditions,” and thus states the hypothesis.

2. Second, scientists look to experience to evaluate the accuracy of their ideas or expectations about the world. That is, they try to find or create the situation that will allow them to observe what they are interested in studying.

3. Through observation and experimentation, scientists can begin to evaluate their ideas and expectations about the world. Learning about the world through observation and experimentation is an example of empiricism, which means nothing more than the acceptance of sensory information as valid.

4. Fourth, on the basis of their observations and experiments, scientists seek to draw conclusions or inferences about their ideas and expectations. They reorganize their ideas and consider the impact of the new information on their theoretical conceptualizations.

Figure 2-2 presents a simplified outline of this procedure, which reflects the evolutionary nature of science. The steps include (1) the development of the hypothesis, (2) the translation of this hypothesis into a research design, (3) the running of the experiment, and (4) the interpretation of the results. You will notice that there is also an arrow from step 4 back to step 1. Researchers take the results and interpretations of their studies and create new research studies that refine the previous hypothesis.

Figure 2-2 Schematic representation of the four major steps in the experimental process. These are developing a hypothesis, designing an experiment, performing the experiment, and interpreting the results.

Figure 2-2 Schematic representation of the four major steps in the experimental process. These are developing a hypothesis, designing an experiment, performing the experiment, and interpreting the results.

In psychological research, we have some powerful techniques to help us achieve this goal. Unlike the detective, who must always reconstruct events after the fact, the researcher has the advantage of creating a new situation in which to test ideas. This is comparable to a homicide detective’s ability to bring a dead man back to life and place him in the presence of each suspect until the murder is reenacted. Such a reenactment might lack suspense, but it would increase the certainty of knowing who committed the murder.

In fact, increased certainty is a large part of the experimental process. Scientists increase certainty by creating an artificial situation—the experiment—in which important factors can be controlled and manipulated. Through control and manipulation, participant variables can be examined in detail, and the influence of one variable on another can be determined with certainty.

Science is a way of determining what we can infer about the world. In its simplest form, the scientific method consists of asking a question about the world and then experiencing the world to determine the answer. When we begin an inquiry, what we already know about our topic leaves us in one of a number of positions. In some cases, we know little about our topic, or our topic may be very complex. Consequently, our ideas and questions are general. For example, how does our memory work? What causes mental illness? What factors make a fruitful marriage? How can we model the brain? Can experience change our brain?

Descriptive Research: What Can We Observe?

The first way to do psychological science is to observe what we see around us. A developmental psychologist might want to know what type of facial expressions newborn babies make. A clinical psychologist could study the specific words that people use to describe distress. A cognitive psychologist might seek to understand how individuals make choices in their lives.

We may want to know how a particular event, such as the beginning of the COVID-19 epidemic, influenced people’s thoughts and emotions. James Pennebaker, Ashwini Ashokkumar, and their colleagues examined the language that people used to track their thoughts and emotions during the initial months of the COVID-19 epidemic. Using a language analysis program, they calculated how much each of hundreds of thousands of posts on Reddit communities used affiliation words such as “we,” “us,” “our,” “together,” and “love,” which they saw as a measure of social relationships. As shown in Figure 2-3, people began to use affiliation words more in the last week of February, indicating that they started to increasingly focus on their social relationships. Mid-March, at about the time when shelter-in-place directives began, there was a sharp increase in feelings of social connection. Overall, COVID-19 increased people’s feelings of social connection.

Figure 2-3 Affiliation words during COVID-19 increased when the COVID-19 crisis began.

Figure 2-3 Affiliation words during COVID-19 increased when the COVID-19 crisis began.

Observation Studies

If little is known about a particular phenomenon, it is often useful simply to watch the phenomenon occur naturally and get a general idea of what is involved in the process. Initially, this is accomplished by observing and describing what occurs. This scientific technique is called observation studies or naturalistic observation.

A classic example of this approach is Charles Darwin’s observation of animals in the Galápagos Islands. His careful descriptions of their appearance and environment became the basis of his theory of evolution. Since the1960s, Jane Goodall has observed primates around the world in their natural environments. She was one of the first researchers to describe a chimpanzee using a stem of grass and modifying it so he could poke it into a termite mound to gain food (Goodall, 1990). Other researchers (De Waal, 2008) sought to understand how animals and humans use peacemaking to resolve aggressive episodes (see Figure 2-4). For example, the bonobo, another primate, uses sexuality in place of aggression. Like humans, bonobos are sexually active at times other than just when a child can be conceived.

Figure 2-4 Chimpanzees. Did you know that in addition to bonobos and humans, chimpanzees often kiss their partner on the mouth after a fight?

Figure 2-4 Chimpanzees. Did you know that in addition to bonobos and humans, chimpanzees often kiss their partner on the mouth after a fight?

How might you perform research with bonobos to determine how they use sexuality for the purpose of lovemaking? It would first be necessary to find a place where bonobos would interact. You would then watch the interaction and note the behaviors observed. It is important that those being observed do not realize they are being observed, otherwise they might change their behavior.

The naturalistic observation method has four characteristics:

1. Noninterference is of prime importance. Scientists using this method must not disrupt the process or flow of events. In this way they can see things as they really are, without influencing the ongoing phenomenon.

2. This method emphasizes the invariants, or patterns, that exist in the world. For example, if you could observe yourself in a noninterfering manner, you might conclude that your moods vary with the time of day, particular weather patterns, or even specific thoughts.

3. This method is most useful when we know little about the subject of our investigation. It is also particularly helpful for understanding the “big picture” by observing a series of events, rather than isolated happenings.

4. The naturalistic method may not shed light on the factors that directly influence the behavior observed. The method provides a description of a phenomenon; it does not answer the question of why it happened.

Case Studies

The naturalistic method may not shed light on the factors that directly influence the behavior observed. The method provides a description of a phenomenon; it does not answer the question of why it happened. Sometimes, as in the case of David and his experience of not recognizing his mother, professionals encounter a rarer event. The study of such a naturally occurring event that happens to an individual is referred to as a case study.

The case study method has a rich tradition in studying unique situations that do not lend themselves to traditional research procedures. In fact, much of our initial understanding of brain function came from careful study of individuals who had accidents or experienced war injuries. The case study offers a means of examining in some depth the manner in which a person understands his or her experiences. Further, the case study offers a means of helping researchers develop new questions to be asked concerning how a phenomenon developed and might be understood.

In psychopathology research, the advantage of the case study is its ability to present the clinical implications of a particular disorder. One classic example is Freud’s case study of Anna O, which focused on the treatment of “hysteria,” which we now call conversion disorder (Breuer & Freud, 1893—1955). Another example is described in Morton Prince’s book The Dissociation of a Personality (1913). Prince described a case of “multiple personality,” what we currently label dissociative identity disorder, at a time when the existence of that diagnostic category was in question. Initial case studies from battlefield experiences also helped to clarify the nature of post-traumatic stress disorder (PTSD).

A recent example of a case study was reported by four health care professionals in China (Yu, Jiang, Sun, & Zhang, 2015). A 24-year-old patient came into their hospital complaining of dizziness and nausea. She also reported that, for more than 20 years, she had experienced difficulty walking in a steady manner. Further examination showed that this individual could speak and understand language, but that she could not stand by herself before the age of 4 and that she did not walk on her own before the age of 7. Brain-imaging techniques showed that this individual was missing an important part of her brain, the cerebellum. The cerebellum is related to motor functions such as walking. Such non-normal cases of development allow scientists to understand how the lack of a cerebellum affects behavior and experience.

Correlational Research: What Goes with What?

Correlational research helps us understand how things go together. As with much of human behavior, there are complex relationships between psychological variables, or factors. Correlational approaches reveal what factors are related. For example, you can ask if peo ple who have more friends have fewer health-related problems than people who do not. Thus, we ask if one aspect of a system is associated with another aspect.

How would you go about answering this question? Let’s begin with the relationship between friends and health. You could start by asking how many Facebook friends an individual has. How does this relate to health? The next step might be to determine the number of times that individual went to the health center. If you did this with a group of people, you would have two numbers for each person—number of Facebook friends and number of health center visits.

Visualizing the Data

What would you do with these data? One helpful technique is to create a visual of the information. A scatterplot is a graph on which the data from each person is plotted. We would use the y-axis to display the number of friends on Facebook (for example, 0—150) and the x-axis to display the number of visits to the health center during the past year (for example, 0—20). We could then look at these measures for each person and plot that point on the graph (Figure 2-5). It is now possible to look at the graph and visually determine if there is a relationship. In general, the dots show a pattern that indicates that people with fewer friends tend to have more health center visits, and the subjects with more friends have fewer visits.

Figure 2-5 Scatter diagram showing negative relationship between two measures. As the number of health care visits increases (y-axis), the number of friends decreases (x-axis). Correlation alone can only describe the relationship between factors. Correlation cannot determine which factor caused the change in the other factor.

Figure 2-5 Scatter diagram showing negative relationship between two measures. As the number of health care visits increases (y-axis), the number of friends decreases (x-axis). Correlation alone can only describe the relationship between factors. Correlation cannot determine which factor caused the change in the other factor.

Determining Direction and Degree

Correlations have both a direction and a degree, or strength. Although humans are good at determining patterns, the correlation coefficient, which indicates both the strength of the relationship and its direction, allows for better precision.

Let’s start with the direction. If the number of friends on Facebook were associated with more health center visits, then this relationship would be called a positive correlation. That is, they both increase. However, if fewer friends were associated with more visits, then this relationship would be called a negative correlation. One increases and the other decreases. Figure 2-6 shows a variety of relationships.

Figure 2-6 Scatter diagrams showing various relationships that differ in degree and direction.

Figure 2-6 Scatter diagrams showing various relationships that differ in degree and direction.

Of course, as you might guess, few factors or variables are perfectly related to one another. Thus, the correlation statistic is also able to reflect the degree of an association between two variables. For example, you might expect that the more hours you study for a test, the higher score you would receive. However, some people who study more will not do as well as some individuals who study less. Thus, it is not a perfect correlation, but a positive one.

Whether the relationship between the variables is positive or negative is denoted by the + and — signs. The correlation statistic can range from — 1 to +1. A perfect positive relationship would be +1 and a perfect negative relationship would be —1. If there was no relationship it would be 0. Thus, a low positive relationship could be.3, whereas a stronger one would be.70. It should be noted that both the statistic and the research approach are referred to as correlational.

Correlation Does Not Equal Causation

In correlational studies, the researcher is interested in asking whether there is an association between two variables, but he or she does not attempt to establish how one variable influences the other, only that a relationship exists. For example, if drinking orange juice at breakfast is associated with better health, it does not mean that the orange juice led to better health. It could be the case that not being in a state of good health would make anyone less likely to want to drink orange juice in the morning.

A correlational relationship does not mean there is a causal one. Establishing that such an association exists may be the first step in dealing with a complex problem and discovering what type of causal relationship exists. In Chapter 6, there is a graph that shows the relationship between the weight of an animal and the number of hours that the animal sleeps (Figure 2-7). If you were to take these data and create a correlation coefficient, the result is a negative correlation of —.8. That is, the more an animal weighs, the fewer hours it sleeps during a 24-hour period. However, this relationship involves only animals that do not eat meat.

Figure 2-7 Length of sleep each day in relation to the weight of the animal. The more an animal weighs, the less it sleeps. The animals shown are ones that do not eat meat. Note a kilogram (1,000 grams) equals 2.2 pounds.

Figure 2-7 Length of sleep each day in relation to the weight of the animal. The more an animal weighs, the less it sleeps. The animals shown are ones that do not eat meat. Note a kilogram (1,000 grams) equals 2.2 pounds.

Source: Siegel (2005).

What we cannot know from correlational research is whether either variable influences the other directly. That is, if two variables are related, what might the reason be for the relationship? It might be that there is a direct relationship. Having lots of friends to follow on Facebook might help you to feel cared for and lead you to feeling less sick. In this way, more friends would result in fewer visits. Looking at the other side of the situation, it could be the case that if a person felt sick often, they might not seek others as friends and post on Facebook.

As you begin to suggest factors that might have produced a high degree of relationship, you realize that a third, unspecified variable actually may have influenced the two variables in the correlational study. For example, there is a high correlation between wearing bathing suits and eating ice cream. In this case, they are both determined by the outside temperature. Thus, a third variable determines the relationship. Thus, the nature of a correlational study is to suggest relationships but not to suggest which variable influences which other variable.

The association of two factors does not in itself imply that one influences the other. However, if there is a low correlation between the two events, you can infer that one event does not cause the other. A high degree of association is always necessary for establishing that one variable influences another. A correlational study is often the first step for providing the needed support for later experimental research, especially in complex areas. At times, correlational research has led to difficult problems for society as you will see in the box: Myths and Misconceptions—Doing Good Science and the Controversy over Childhood Vaccinations and Autism Spectrum Disorder.

Myths and Misconceptions—Doing Good Science and the Controversy over Childhood Vaccinations and Autism Spectrum Disorder

Over the past 20 years, our understanding of autism spectrum disorder has increased tremendously. Autism is a disorder in which children show social difficulties in terms of communication with others. Parents and guardians are paying more attention to the social functioning of their children and even seeking evaluations to identify autism early in a child’s life. What are the consequences of this? Given that more children are being evaluated, one consequence of this is that more children are being diagnosed with the disorder. Is there an increase in the disorder? That is an important question.

During the past 20 years, many people in the United States also became concerned with the food they ate. Sales of organic food increased. Now, what if we presented a graph of organic food sales and the diagnosis of autism? As you can see in Figure 2-8, there is a close correlation.

Figure 2-8 This graph shows a close positive correlation (r=.99) between the sale of organic foods and the diagnosis of autism over 20 years.

Figure 2-8 This graph shows a close positive correlation (r=.99) between the sale of organic foods and the diagnosis of autism over 20 years.

You might conclude that eating organic food causes autism. However, you are a better scientist than that! Since the correlation between organic food sales and autism diagnosis was high, you would want to look for the reason for the relationship. You could start by researching if those children who ate organic food developed autism. To answer this question, you could note the number of children who ate organic food and developed autism and the number who did not. You would also want to see the number of children who developed autism but did not eat organic food. You might examine if mothers if who ate organic food during their pregnancy had children who later developed autism. You could also look for third variables that were related to both. For example, as society became more interested in health, then both eating better and seeking an early health diagnosis for a disorder such as autism might increase.

As you realize, making the conclusion that eating organic food leads to autism is bad science. However, there are times that fraud and not science leads to society believing in an erroneous conclusion. In 1998, a British surgeon and medical researcher, Dr. Andrew Wakefield, published a paper in the respected medical journal The Lancet. The article, based on 12 children, reported a relationship between the measles, mumps, and rubella (MMR) vaccine and autism and a certain bowel disease. The paper suggested that 8 of the 12 children might have shown problems within days of being given the MMR vaccine. As you can imagine, this published report upset many parents, some of whom refused to have their children vaccinated.

Those in the medical profession found the claim of the relationship between the MMR vaccine and autism troubling and began to research this relationship. In studies in other labs, no one was able to find the same relationship. Media investigations discovered that Andrew Wakefield had been given money by a lawyer interested in suing drug companies on behalf of children with autism. The British General Medical Council opened an investigation and concluded that the published report involved dishonesty. Andrew Wakefield lost his medical license. Medical journals and other publications referred to the published paper and other presentations of Wakefield as fraudulent. The Lancet retracted the paper.

Yet the initial media discussion of Wakefield’s article left a profound “hangover” with chilling effects. In the wake of these allegations, vaccination rates dropped in both England and the United States. As recently as 2015, an outbreak of measles was traced to unvaccinated children visiting Disneyland in California. In 2019, the US had the greatest number of measles cases reported since 1992 and since measles was declared eliminated in 2000.

Thought Question: What are your thoughts concerning the following question for psychology and society at large: Why are people so quick to believe in the initial claim of an MMR vaccination and autism connection and so slow to give up this belief in the presence of overwhelming data?

Experimental Research: How Do We Determine Cause and Effect?

Establishing that a relationship exists between two events does not allow us to determine exactly what that relationship is, much less to determine that one event actually caused the other event to happen. If we want to state that one event produced another event, we need to develop a much stronger case for our position.

To do this, we could see how some single event over which we have control affects the phenomenon we are studying. In this way, we begin to interact with the phenomenon. We structure our question in this form: “If I do this, what will happen?” Numerous questions can be asked in this way, such as, “Will you learn words better in a foreign language if each word is of the same class (for example, food words) instead of coming from different classes (foods, cars, and toys)?”

As our knowledge grows, we may even get to the point of formulating specific predictions. In this case, our questions are structured in the form, “If I do this, I expect this will happen.” Sometimes our predictions are more global, and we predict that one factor will be stronger than another. We might predict that more people are likely to help a stranger if they perceive the environment to be safe than if they think it is dangerous. Sometimes, however, we may know enough about an area to make a more precise prediction, or point prediction. For example, we might predict that three months of exercise will lead to a 10-mm Hg decrease in blood pressure. These approaches, in which we interact directly with the phenomenon we are studying, are examples of the experimental method.

The Experimental Method

What if we want to know whether drinking coffee affected how many words from a list could be remembered? The hypothesis, or idea being tested, is that drinking coffee influenced memory. To test the hypothesis, we could give coffee to one group and not to a second group. The group that received the coffee is called the experimental group. The group that did not receive the coffee is called the control group. A control group is a group that is treated exactly like the experimental group except for the factor being studied. In this case, the factor being studied is the coffee and its influence on memory.

Since memory can be viewed in a variety of ways, we need a specific definition of what memory means in our study. Memory, for example, could be defined as the number of words that could be remembered from a list of 20 words. An operational definition takes a general concept, such as memory, or aggression or effectiveness, and places it within a given context. That is, it redefines the concept in terms of clearly observable operations that anyone can see and repeat. For example, we might define aggression as the number of times a child hits a toy after watching a violent television show.

In an experimental study, we want to know how one variable—which we manipulate—affects another variable. In the example of coffee affecting memory, whether or not someone drank coffee was the manipulated variable. This is also called the independent variable (IV). Memory, in this example, is the variable influenced by the coffee and is called the dependent variable (DV). That is, it depends on, or is influenced by, the independent variable.

What other factors could influence a memory test? If we suspect that some unintended factor may also be operating, then the truth or validity of the experiment is seriously threatened. Thus, the conclusion that the IV influenced the DV could be questioned.

In the memory experiment, if the control group was run in the morning and the experimental group in the afternoon, then time of day could have an effect. Whenever two or more independent variables are operating, the unintended independent variables (those not chosen by the experimenter) are called confounding variables. Other confounding variables may go together with the independent variable and be more difficult to notice. For example, assume that a researcher compared a new medication against a problem-solving approach for the treatment of anxiety. If she found the problem-solving approach to show a greater reduction in anxiety, could she conclude that problem solving produced the reduction?

Although that is one possibility, it also may have been the case that spending time with a professional produced the reduction in anxiety. That is, because giving medications requires less time with a client than discussing problem-solving techniques, the results found may not have been due to the independent variable as planned in the study but rather to the confounding variable of time spent with the patient.

Combination Research Designs: When Does It Make Sense?

Let’s examine a specific study in which both experimental and correlational procedures were used. Thomas Elbert and his colleagues began with the idea that experience could change the manner in which connections in the brain were established (Elbert et al., 1995). These researchers needed to find a task that would allow them to measure change and make logical inferences. Since they were interested in long-term changes, they focused on the skill of playing a musical instrument, something people often learn in childhood. Specifically, they chose the violin. Here is where logic and experimental design came in. By focusing on the violin, these researchers were able to compare the differences between the violinists’ left and right hands and their representation in the brain.

Violinists use their left hand to continuously finger the strings. The right hand simply moves the bow back and forth. If playing a violin for 20 years would affect the brain, then it should affect those areas involved with the left hand in a different manner than those involved with the right. This is exactly what they found.

To measure neuronal activity in the brain, these researchers used a brain-imaging technique, the magnetoencephalography (MEG), which measures magnetic activity produced in different parts of the brain. In their study, they found that neuronal activity in the brain was different between the areas of the brain related to musicians’ left and right fingers. Further, they found that the brain areas of the musicians’ right hands were not different from those of the control group who did not play a musical instrument.

Thus, the experimental comparison was between individuals who had played a musical instrument since childhood and the control group was those who had not. There was also a comparison between the brain areas involved with the left and right fingers of the musicians. Further, these researchers examined the correlation between neuronal activity (dipole strength) and the length of time an individual had played an instrument. Figure 2-9 shows this relationship. As you can see, this is a negative correlation in that the earlier (lower number) one began to play an instrument, the stronger the neuronal activity was. The experimental and correlational aspects of this research helped the researchers logically conclude that previous experience can influence brain organization. We have not always thought that actions could change our brains, as you will see in the box: The World Is Your Laboratory—Science from a Larger Perspective.

Figure 2-9 Relationship between neuronal activity and the length of time an individual had played an instrument. This graph shows that those who began playing the violin at an early age showed a different pattern of brain activity than those who began playing later or never played the violin.

Figure 2-9 Relationship between neuronal activity and the length of time an individual had played an instrument. This graph shows that those who began playing the violin at an early age showed a different pattern of brain activity than those who began playing later or never played the violin.

Source: Elbert, Pantev, Wienbruch, Rockstroh, and Taub (1995).

The World Is Your Laboratory—Science from a Larger Perspective

Science begins with someone saying, “There is something I don’t understand and I want to understand it.” The task then becomes one of figuring out how to gain this understanding. Sometimes, it is by watching; sometimes it is by interacting with the process. In this sense, science is universal since we, as humans, seek to gain knowledge and understand ourselves and our world.

Although science seeks to add new information, it is also influenced by scientists’ own human nature. As humans we live in a culture that influences what we value and how we interact with others. Thomas Kuhn (1970) has examined the larger culture of science. From this perspective, he suggests that science is more than just adding new information to old facts. For Kuhn, over time scientific information reflects a cycle of revolutions followed by periods in which normal science takes place. The period of normal science reflects what Kuhn called a paradigm.

A paradigm is a set of assumptions that guides the activity until a new revolution (a paradigm shift) takes place. If you lived in the Middle Ages, when it was assumed that the world was flat and the center of the universe, then all of your activities would reflect this “fact.” You would draw maps that showed the world to be flat and never question it. However, with the realization that the world was neither flat nor the center of the universe, your view of the world would change, a paradigm shift.

Psychology and the neurosciences have gone through a number of these revolutions. At the beginning of the 20th century, it was assumed that everything was learned and the infant came into the world as a blank slate. Research today reflects a complex interplay between the environment and genetic processes. For example, we now know that children come into the world ready to absorb the languages available to them. Likewise, it was previously assumed that once you left childhood, your brain no longer could be changed. We now know that everything you do has the potential to make changes in the brain. In short, with new scientific discoveries, paradigms shift.

Thought Question: What “scientific fact” is so taken for granted as true that it might just be ready for a new way of looking at it through a paradigm shift?


✵ Define the following: science, facts, scientific knowledge, doubt, and falsification. How do they fit together?

✵ What is the difference between science and pseudoscience? What criteria would you use to decide for yourself whether a particular report is science or pseudoscience?

✵ What are the four major steps in the experimentation process?

✵ What is the primary value of a case study? In what situations in psychology would it be most appropriate?

✵ What are the four characteristics of the naturalistic observation method?

✵ If a research study reports that two variables are correlated, what do you know about their relationship? What don’t you know about their relationship?

✵ How is the experimental method different from naturalistic observation and the correlational approach?

✵ What is the difference between the experimental group and the control group? Why is the control group important?

✵ What is an operational definition and why is it important in doing research? Create at least two different operational definitions of “violence on television” and “aggression.”

✵ How are the independent variable (IV), the dependent variable (DV), and confounding variables related?

✵ Thomas Kuhn’s “culture of science” includes revolutions, normal science, and paradigms. What are they and how do they work together?

How Do Scientists Evaluate and Analyze Data?

Perhaps you have heard the story of our friend from Boston who, every morning, went outside, walked around in a circle three times, and yelled at the top of his voice. His neighbor, being somewhat curious after days of this ritual, asked for the purpose behind his strange behavior. The man answered that the purpose was to keep away tigers. “But,” the neighbor replied, “there are no tigers within thousands of miles of here.” To which our friend replied, “Works quite well, doesn’t it?”

How could we demonstrate to our friend that his yelling is not causally related to the absence of tigers? One strategy might be to point out that the absence of tigers is due to other reasons, including the fact that there are no tigers roaming the greater Boston area. In technical terms, we would say that yelling could be a necessary condition but not a sufficient condition for the absence of tigers.

Logic is particularly important in science when asking, “What question should my experimental study answer to test my ideas about the world?” For example, our friend’s reasoning was incorrect because it overlooked many other plausible explanations for the obvious absence of tigers. Although our friend sought to infer a relationship between his yelling and the absence of tigers, his inference was weak. Inference is the process by which we look at the evidence available to us and then use our powers of reasoning to reach a logical conclusion. Like Sherlock Holmes engaged in solving a mystery, we attempt to solve a problem based on the available evidence. Did the butler do it? No, the butler could not have done it because there was blond hair on the knife and the butler had black hair. But perhaps the butler left the blond hair there to fool us.

Like a detective, scientists try to determine other factors that may be responsible for the outcome of their experiments or to piece together available information and draw general conclusions about the world. Also, like the detective, the scientist is constantly asking, “Given these clues, what inference can I make, and is the inference valid?” Logic is one method for answering these questions.


Logical procedures also help us understand the accuracy or validity of our ideas and research. Validity means that something is true and capable of being supported. There are various types of validity in psychology that arise from differing contexts. These contexts range from developing types of tests to running experiments. The overall question is, “Does a certain procedure, whether it is an intelligence test or an experiment, do what it was intended to do?” There are two general types of validity (Campbell & Stanley, 1963).

Internal validity. The word “internal” refers to the experiment itself. Internal validity asks, “Is there another reason that might explain the outcome of our experimental procedures?” Students are particularly sensitive to questions of internal validity. For example, when it is time for final exams; they can make a number of alternative suggestions about what the exam actually measures and why it does not measure their knowledge of a particular subject. Like students, scientists look for reasons (threats to internal validity) that a particular piece of research may not measure what it claims to measure. In the case of our friend from Boston, the absence of tigers near his house could have reflected a long-standing absence of tigers in his part of the world rather than the effectiveness of his yelling.

External validity. The word “external” refers to the world outside the setting in which the experiment was performed. External validity often is called generalizability. External validity refers to the possibility of applying the results from an internally valid experiment to other situations and other research participants. Thomas Elbert and his colleagues found brain differences between those who played an instrument from an early age and those who did not. External validity asks if these results could be found in other individuals who played musical instruments from an early age.

We logically design our research to rule out as many alternative interpretations of our findings as possible and to ensure that any new facts are applicable to as wide a variety of other situations as possible. In many real-life situations in which external validity is high, however, it is impossible to rule out alternative interpretations of our findings. In a similar way, in laboratory settings in which internal validity is high, the setting is often artificial, and in many cases our findings cannot be generalized beyond the laboratory. Consequently, designing and conducting research is always a trade-off between internal and external validity. Which type of validity we choose to emphasize depends on the particular research question being asked.


Whereas validity refers to a measurement being true, reliability refers to the measurement being consistent. For example, a bathroom scale is reliable if it gives you the same weight no matter how many times you stand on it (assuming your weight does not change between weighings). Most of our measures such as intelligence or treatment effectiveness show more variation than a physical measure such as weight. For this reason, when researchers are first designing a measure for a particular construct (for example, depression), it is common for the researchers to give the same measure over a number of occasions under similar conditions. They then examine the consistency of the research participants’ responses to determine test—retest reliability, which is the correlation between the scores on each of the testing occasions.

In any study you want to use instruments that show a high positive correlation between these occasions. In addition, we also need the people who make the observations in a research study to be reliable in their ratings of particular behaviors. Like test—retest reliability, we can correlate the ratings of different observers when examining the same behavioral pattern. This is referred to as interrater reliability.

The Importance of Replication

Before we continue, let’s clear up one misconception. It is the idea of designing “the one perfect study.” Although scientists strive to design good research, there are always alternative explanations and conditions not included in any single study. It is for this reason that Donald Campbell, who introduced scientists to the idea of internal and external validity, also emphasized the importance of replicating studies (Campbell, 1957).

Replication involves repeating the study. If the same study is performed a number of times with similar results, then we can have more assurance that the results were both reliable and valid. Even better, if the study is performed in a variety of settings around the world, we have even more confidence in our results (Koole & Lakens, 2012; National Academy of Sciences, 2019; Shrout & Rodgers, 2018). The American Psychological Society (APS) is supporting such groups as the Center for Open Science ( in getting researchers to register their replication study before it is performed. In this way one can see after a number of the replication studies are completed which ones found data consistent with the original study and which did not. Thus, as an informed consumer of research findings, you want to base your conclusions on findings that have been replicated. There are also statistical techniques that you will learn about in more advanced courses that allow us to combine the data from a number of separate studies examining the same variables. This is referred to as meta-analysis. Meta-analysis also lets us know the strength of the relationship between the independent variable and dependent variable.


Behavior can be described in many ways. In naturalistic observation studies, descriptions may consist of simple lists of behaviors or behavior sequences. For example, a cognitive psychologist might describe how a person interacts with a computer app. A developmental or clinical psychologist might report the manner in which a family interacts. In experimental studies, behaviors usually are expressed in more quantitative terms. We need a measure that others can use also. In a diet study, a common measure would be weight in pounds. Some other examples of quantitative measurements are the number of millimeters by which an observer overestimates the length of a line in a visual illusion experiment, the number of words recalled in a memory experiment, the change in a person’s heart rate as a result of watching an emotionally charged film, and the results of an IQ test taken under different conditions.


There are a variety of ways to measure a particular process. If you were interested in emotional responding, for example, you could record facial expression; measure psychophysiological activity, facial muscle activity, brain activity, and heart rate; obtain self-reports of internal states; and so on. Recording more than one measure is of great benefit to understanding what you are studying. Even when you decide which measure or measures to obtain, you still must decide whether to record the frequency of response, the intensity of response, the duration of response, the reaction time to the first response, or some combination of measures most appropriate to the question being asked. But how do we determine the appropriateness of a measure? One answer to this question is related to the conceptual question being asked.

Graphic Description of Frequency Information

A useful way to begin analyzing the results of any experiment is to convert your numerical data to pictorial form and then simply look at them. One way to do this is to draw a frequency distribution. In a frequency distribution, you simply plot how frequently each score appears in your data. Suppose you were interested in dreaming. An initial baseline measurement might be to ask 20 people to write down their dreams for a week. To get an idea of how often people dream, you might begin your analysis by seeing how many dreams each person recalls.

To create a pictorial representation, you want to match each participant with the number of dreams they recalled. A first step would be to find the smallest number of dreams recalled (0 in our case) and the largest number of dreams recalled (7). You could then list all the numbers from 0 to 7. Going through all of the responses recorded in Table 2-1, you could make a mark by the number of dreams recalled for each person in your study.

Table 2-1 Table of data from hypothetical dream study. This shows the number of dreams that each of the 20 participants reported.

Hypothetical Dream Study


Number of Dreams Recalled


Number of Dreams Recalled









































Now we have two variables (number of dreams recalled and number of people who recalled a certain number of dreams). This information can be plotted on a graph. Figure 2-10 is a frequency distribution for these data. The vertical, or y-axis (ordinate), labeled “Frequency,” is the number of people who fall into each category; the horizontal, or x-axis (abscissa), labeled “Score,” is the number of dreams recalled. This type of presentation is called a bar graph.

Figure 2-10 Bar graph of dream data. This graph shows the number of individuals who reported a particular number of dreams.

Figure 2-10 Bar graph of dream data. This graph shows the number of individuals who reported a particular number of dreams.

Descriptive Statistics

Descriptive statistics are statistics that define the nature of the number that we are examining. You use descriptive statistics all the time in terms of your weight, your grades, and the amount of money you spend. Descriptive statistics allow us to characterize numbers in terms of frequency and variability.

Measures of Central Tendency

If you pay attention to the news, you will notice that certain terms are often used to describe the way we “typically” live. You will hear one report discussing the median income of college professors; another report may discuss the mean price of a new house in different cities in America. Modal (mode) descriptions are used less often, which is simply the most frequently occurring number.

There are three measures of central tendency, which is simply a single summary measure that describes a number of scores. These central tendency measures are known as the mean, median, and mode. Of these three, the mean is used most often. The mean of a set of scores is the arithmetic average of those scores. It is obtained by adding the scores and dividing the total by the number of scores. In our dream data, 20 people reported a total of 70 recalled dreams:

In this particular example, 70 divided by 20 equals exactly 3.75.

The median of a set of scores is the middle score—that is, the score that has an equal number of scores both above and below it. To calculate the median of a set of scores, list all the scores in order of magnitude (from largest to smallest or vice versa); the median is the middle score or, in the case of an even number of scores, the score halfway between the two middle scores. In our dream data, the two middle scores are 3 and 4. Consequently, the median is 3.5.

The mode is the most frequently occurring score. The only mathematical calculation required to compute the mode of a distribution of scores is to count the frequency of each score. The score that occurs most frequently is the mode. If there are two scores with the same frequency, we report two modes. For example, the dream data presented earlier had two modes: one at two dreams and one at five dreams recalled.

Given these three measures, you may be wondering which provides the best estimate of central tendency. There is no clear-cut answer. The appropriate choice varies with the particular frequency distribution and the intent of the researcher. For example, in a normal distribution the mode, median, and mean all have the same value (Figure 2-11a). This is not the case in a skewed distribution (Figure 2-11b). In a skewed distribution, extreme scores affect the mean.

Figure 2-11 Mean, median, and mode of (a) a normal distribution and (b) a skewed distribution. In a normal distribution, the mean, median, and mode are the same. If the distribution is not normal, then the mean, median, and mode can be different.

Figure 2-11 Mean, median, and mode of (a) a normal distribution and (b) a skewed distribution. In a normal distribution, the mean, median, and mode are the same. If the distribution is not normal, then the mean, median, and mode can be different.

Conceptually, the question asked influences which central tendency measure is used. If you are interested in the scores of a whole group, then the mean is the most appropriate measure. However, if you are more interested in a representative individual score, then the median is more appropriate. For example, before you take a job in a company where the mean salary is $100,000 a year, you might want to find out the median salary. That is, a mean of $100,000 could be produced by 10 people making $10,000 each and 1 person making $1,000,000 (1,100,000/11 = 100,000). Thus, mean salary would not help. If you were a stock market analyst concerned with how much a company pays out in salaries, however, the mean would be the appropriate measure. In other words, if you are interested in the scores of a whole group, then the mean is the most appropriate measure. However, if you are more interested in a representative individual score, then the median is more appropriate.

In psychological research, we tend to use the mean most often both for historical reasons and because it fits into already developed statistical theory. In summary, the measure of central tendency that you use depends on the question that you are asking. In the final analysis, you must use your judgment to determine which measure of central tendency to use.

Measures of Variability

You probably have completed a science course assignment requiring you to make measurements of some physical dimension, such as the length of a line. The exact readings among students probably differed. If we were to plot each measurement, we would see that the measurements varied around a particular point. We could also imagine that contained within these measurements was the actual length of the line or the true score. Any measurement is composed of a true measurement plus the variability associated with that measurement.

For example, your weight may change from day to day while still remaining constant when viewed over a series of measurements. Thus, to describe a set of measurements more accurately, we need a different measure. As we have seen, a measure of central tendency gives us some information concerning a set of scores. However, it does not give any information about how the scores are distributed. To obtain a more complete description of a set of data, we use a second measure in addition to the measure of central tendency. This is a measure of variability.

Measures of variability are attempts to indicate the degree to which scores in a data set differ from one another. We can also call this variability dispersion. One common measure of dispersion is the range, which reflects the difference between the largest and smallest scores in a set of data. The actual computational formula given for the range in most introductory statistics books is the largest score minus the smallest score. (In the dream data presented earlier, the range is 7; that is, 7 — 0 = 7.) Although computing the range is easy, the range tells us only about the two extreme scores; it provides no information about the dispersion of the remaining scores if we know nothing about the underlying distribution.

Table 2-2 Data table of two different groups.

Group A

Group B



















Let’s now demonstrate this graphically. Consider the following two sets of data for Group A and Group B.

In these two distributions, the ranges are the same (8), and the means are the same (6), yet the actual shapes of the distributions are very different (Figure 2-12). The scores in Group B appear more concentrated in the center of the distribution, yet our estimate of range does not reflect this. To provide a more sensitive description of the dispersion of all the scores, we use a second measure of the variability of data. This measure is called the variance. The variance is a description of how much each score varies from the mean.

Figure 2-12 Two different distributions with the same range and mean but different dispersions of scores. To fully understand the numbers in your experiment, you would want to consider the distribution of scores.

Figure 2-12 Two different distributions with the same range and mean but different dispersions of scores. To fully understand the numbers in your experiment, you would want to consider the distribution of scores.

Another measure of variance is standard deviation. A standard deviation is defined mathematically. It is determined by taking the square root of the variance. The standard deviation is the most common measure of dispersion used in the scientific literature. By knowing the standard deviation, you can know how the scores in your study contribute to the mean. That is, you can know if they fall close to the mean or are widely dispersed.


✵ Identify and describe the two general types of validity. What is the importance of each in conducting good research?

✵ What are three measures of central tendency? How is each calculated, and when would you use it?

✵ What are three measures of variability? How is each calculated, and when would you use it?

How Do We Design an Experimental Study?

The goal of experimental research is to determine the relationship between the independent variable and the dependent variable. The experimental approach is all about strong inference. The less bias there is in terms of demand characteristics related to both the participant and the experimenter aids in creating a logical relationship between the IV and DV. How do we infer a relationship between the independent variable and the dependent variable?

There are four factors that are critical to sound inference (Figure 2-13):

1. Participant selection

2. Participant assignment

3. Design of experiment

4. Interpretation of relationship of IV to DV

Figure 2-13 The four conceptual steps in experimentation. The first is selecting the people to be in the study. The second is assigning the participant to groups such as the experimental and control group. The third is to perform the experiment. The fourth is to interpret the results.

Figure 2-13 The four conceptual steps in experimentation. The first is selecting the people to be in the study. The second is assigning the participant to groups such as the experimental and control group. The third is to perform the experiment. The fourth is to interpret the results.

Select and Group Participants

Selecting participants for an experiment may seem simple, but it is often a bigger problem than it appears. The individuals selected for the study should be directly related to the hypothesis being tested. If you want to know if college students perform better on a memory test after drinking coffee, then of course you select college students. If you want to be able to apply the results to the population of all adults, then it is necessary to include adults of all ages in the study. Each study begins by asking who the individuals are that this research will apply to. This is formally known as the population under study. However, in most cases you can’t study everyone.

Researchers use a variety of means for selecting participants. The best is some form of random sampling. Random sampling is the case in which any person that is part of the population is equally likely to be selected. You might choose every tenth person in the college directory, for example. When the selection is not random, that is, every person is not equally likely to be chosen, then bias can appear. One classic example of this type of bias happened in the surveys of the 1936 presidential election.

In 1936, asking more than 2 million people for whom they would vote in the next presidential election led one magazine to conclude that the Republican nominee, Alfred Landon, would beat the Democratic incumbent, Franklin Roosevelt (Gallup, 1972). To be precise, the magazine predicted that Landon would carry 32 states and 57% of the vote compared to Roosevelt’s 43%. In fact, the results were very different, with Landon carrying only 2 states and receiving 37% of the vote in comparison to Roosevelt’s 61%.

George Gallup conducted a much smaller survey of only 5,000 voters and correctly predicted the winner. What was the secret of his success? How did Gallup do more with less? The answer was in the sampling procedures. Although the magazine sampled 2,376,523 people, their names were drawn from automobile registration lists and telephone directories.

Realizing that this survey took place in 1936, during the middle of the Great Depression, you can understand, perhaps, what the bias was in this sampling technique. During the Depression, many people did not have money for cars or telephones and thus would not have been included in the survey. Those who did have telephones and cars tended to be wealthy and also to be Republicans. Thus, wealthy Republicans completed the survey in disproportionate numbers, saying they would vote for Landon and against Roosevelt. Gallup used a sampling procedure that did not exclude the poor, who were mainly Democrats and more likely to vote for Roosevelt. Gallup thus had the more accurate representation, and his prediction of a Roosevelt victory was correct.

After the participants have been randomly selected from the larger population under study, they can then be randomly assigned to experimental and control groups. This ensures that the groups are equal before the experiment begins. Randomization controls for both known and unknown potentially confounding variables. Randomization leaves solely to chance the assignment of our participants to a group. In this way any differences, even unknown or unsuspected differences, will also be nullified by being randomly distributed between our two groups.

Plan the Structure

Somewhat like a blueprint, the experimental design directs the procedures and gives form to the experiment. In essence, an experimental design is a plan for how a study is to be structured. In an outline form, a design tells us what will be done to whom and when. To be evaluated favorably, a design must perform two related functions. First, it must provide a logical structure that enables us to pinpoint the effects of the independent variable on the dependent variable and thus answer our research questions. Second, it must help us to rule out confounds as an alternative explanation for our findings. A sound design must allow us to determine logically the effect of the independent variable on the dependent variable and to rule out alternative explanations.

Imagine a study in which a social psychologist was interested in determining if watching a film about the value of interacting with people (IV) would result in the participant being more sensitive to individuals showing more positive or engaging emotions (DV). After the subject watched the film, he or she could be shown images of a variety of individuals with different facial expressions. The participant could be instructed to press the computer key as soon as he or she could identify the emotion.

If we were to diagram the design of this study, it would be:

Select the group

Watch film

Measure reaction time

If we performed the experiment with just a single group, what could we conclude? We could determine if our participants were faster recognizing positive or negative emotions. However, this would not help us determine if this was related to the film they saw. It may be that humans recognize positive versus negative emotions differently no matter what the situation. Such a design would not be much help in pinpointing the effect of the independent variable on the dependent variable, nor would it rule out confounds.

A stronger design would use a control group. This design would appear as:

Experimental group


Measure reaction time

Control group

No film

Measure reaction time

Since the control group had not seen the film, then we would have stronger evidence that seeing the film was related to differences in reaction time measures.

In our research design, different participants comprise the experimental group versus the control group. This type of design is called the between-subjects design. Between-subjects refers to the fact that different participants are part of each group.

Interpret the Results

Once we have collected our data we want to know how to interpret our experimental results. We do this by considering three separate hypotheses:

1. Null hypothesis—this is a statistical hypothesis that is tested to determine if there are differences between our experimental and control groups. Null is similar to being invalid or zero and thus refers to the possibility that there are no differences between groups. Part of this statistical procedure is to ask the question of whether our results could have happened by chance.

2. Confound hypothesis—this is a conceptual question that asks if our results could be the result of a factor other than the independent variable.

3. Research hypothesis—this asks the question of whether our results are related to the independent variable.

Inferential Statistics

As described previously, descriptive statistics are used to determine characteristics of a group such as mean and variance. You might want to know how many other children 6-year-olds talked to during recess. You might examine how students performed on a particular college quiz. In both of these cases, you are only interested in these specific groups. However, we usually perform research with a limited number of individuals so that we can infer the behavior of all individuals related to the group we are studying. For example, if we study how a group of 3-year-olds produce sentences, then it is our assumption that other 3-year-olds would perform in a similar way.

What are the odds that the 3-year-olds in our group are like all 3-year-olds everywhere? To determine this probability, we use inferential statistics, which are methods for understanding how the data from an experiment can be generalized to all similar people. Ideally, if the same experiment was run an infinite number of times with a different sample of individuals chosen from the entire population, then the population of estimates would then represent all the possible outcomes of the experiment.

What we need, of course, is a technique for determining whether a set of results is different from what would be expected. One of the common statistical techniques used for this is called the t-test. A t-test determines the size of the association between two variables in relation to the number of participants in the study. It was developed near the beginning of the 20th century by William Gosset who worked for the Guinness Brewery in Dublin, Ireland. Since breweries want all of the beer they sell to taste the same, Gosset wanted a method for determining the uniformity of each batch of beer. In this case, he actually wanted the null hypothesis to be true. That is, each beer would be the same.

Confound Hypothesis

The second question asked is could our results be due to a confound, something that systematically biases the results of our research, rather than the independent variable. What is a confound? Almost anything can be a confound. Confounds often result from unintended factors. It may be the fact that on the day before we ran the experimental group in a social psychology experiment on conformity there was a television special on how we always do what other people tell us. Or a confound may be introduced into a weight-reduction experiment if one group is asked to lose weight at a time of the year when people eat less and their gastrointestinal systems work faster (summer), whereas another group was started at a time when people eat more (winter). Some confounds can be prevented or controlled, but others cannot. You cannot control world events, but you can ask whether there is any reason to believe that a particular event that took place inside or outside of the laboratory could have influenced one group more than another and thus introduced a confound.

One classic example of an initially overlooked participant factor occurred in an experiment conducted in the 1930s at the Hawthorne plant of Western Electric. It was designed to determine how factors such as lighting and working hours affect productivity (Mayo, 1933). The participants were a group of women who worked at the plant; they were asked to work under varying conditions. The productivity of these women was compared with general productivity. When the data were analyzed, a strange finding emerged. In many cases the productivity of these women increased. It even increased under a condition in which the lighting in the experimental condition was not as good as that in the actual plant. From an experimental design standpoint, it was difficult to understand the data until the experimenters examined how they had treated these women. The answer they came up with was that the women in the experiment had been given special attention just by being in an experiment. This attention caused the women to consider themselves special, and this feeling was reflected in the work they performed. This came to be known as the Hawthorne effect after the name of the plant in which they worked. In research terminology, this type of bias is referred to as demand characteristics. Demand characteristics occur when a participant’s response is influenced more by the research setting than by the independent variable.

A related phenomenon is the placebo effect. This occurs when some people show psychological and physiological changes just from the suggestion that a change will take place. The placebo effect is related to prior learning, expectation, and social factors, and also includes a brain component (Wager & Atlas, 2015; Geuter, Koban, & Wager, 2017). It includes both situational factors and what the person tells himself or herself (see Figure 2-14). How could this effect be addressed in a treatment study to reduce anxiety?

Figure 2-14 The placebo response includes both external and internal factors. External factors can include where the procedure takes place, how the professional appears, and the nature of the social interaction. Internal factors can include expectations, past experiences, and internal reactions to the professional.

Figure 2-14 The placebo response includes both external and internal factors. External factors can include where the procedure takes place, how the professional appears, and the nature of the social interaction. Internal factors can include expectations, past experiences, and internal reactions to the professional.

Source: Wager and Atlas (2015).

To control for the placebo effect in research, various procedures have been used. One is to use a control group that receives either no treatment or a treatment previously shown to be ineffective for the particular disorder under study. In medical research, it is common to give a “sugar” pill that looks exactly like the medicine with the active ingredient.

Another procedure to study placebo effects is called the open-hidden paradigm (Geuter, Koban, & Wager, 2017). For example, after an operation in a hospital, patients are often connected through a needle in their arm to a machine that delivers various drugs. Since the machine is there constantly, most individuals don’t actually know when medication is being delivered. This is the hidden part of the open—hidden paradigm. The open part of the paradigm is for the health care worker to administer the drug in the form of a shot or to place the drug in the machine so that the patient realizes that it is being administered.

With pain patients, open administration has been found to be more effective in reducing pain than hidden administration (Price, Finniss, & Benedetti, 2008; Kaptchuk & Miller, 2018). In one of these studies, which used five different drugs including morphine, pain patients showed more pain reduction when the same drug was administered in the open as opposed to the hidden condition (Amanzio, Pollo, Maggi, & Benedetti, 2001).

During the 1970s, researchers were trying to understand how opiate drugs such as heroin, morphine, and opium worked. To their surprise, they found that there are opiate receptors in our brains. This suggested that our brains use opiate substances for some purpose. To make a long story short, they discovered that not only did the brain have receptors for opiates but also that the brain made opiates. The most well-known of these brain-made opiates are called endorphins. The term endorphin is a contraction for “endogenous morphine.” Like the opiate drugs given by medical professionals for pain, our brain administers endorphins as a way to reduce pain. This helps us to understand how athletes or solders are able to complete an activity even when injured. At times, they don’t even know they are injured.

This research with endorphins may lead one to believe that our brain is involved in the placebo effect. However, how do we know that? One way we know that begins with a study from 1978 (Levine et al., 1978). In this study, individuals who had seen the dentist for oral surgery were given medication for pain two hours later. Some of these dental patients were given morphine, some were given a placebo, and some were given a placebo plus naloxone. Naloxone is a chemical that takes away the effects of morphine. Both the morphine and the placebo reduced the pain. However, the group that received naloxone showed an increase in pain. This was clear evidence that the placebo effect for pain involves the opiate substances produced by our brain.

Experimenters also have expectations. For example, knowing that one set of participants has been assigned to one condition rather than another could result in those participants being treated differently. Such situations are referred to as experimenter effects. One important way to control for experimenter effects is to use a double-blind experiment in which the experimental group is divided into two groups. One group is given the actual treatment, and the other is given a treatment exactly like the experimental treatment but without the active ingredient. Neither the placebo group nor the experimental group would know which medication they are receiving, and in this way these research participants are said to be blind controls. The term double blind indicates that the experimenters giving the medication also do not know which treatment is experimental and which is placebo.

In order to sort through the results of our experiments, we must be like detectives who constantly ask if there is another way to understand what was found. Our way of doing that is through research, logic, and doubt. We use research to design a study to consider alternative possibilities. We use logic to consider if our conclusions follow from the results. We use doubt to ask if there is a way to know if we are wrong.


✵ What are four factors critical to enabling sound inference in determining the relation between the IV and the DV in an experiment?

✵ Why is randomization important to selecting participants and assigning them to groups in an experimental study?

✵ What is the difference between descriptive statistics and inferential statistics? What is the primary purpose of each?

✵ What is a double-blind experiment, and what are researchers trying to control for by using that design?

What Are the Ethics of Scientific Research?

In several Nazi concentration camps during World War II, prisoners were injected with a virus or bacterium and then received drugs to determine the drugs’ effectiveness against the injections. Although medical knowledge was gained from these experiments, the world judged the experiments unethical and criminal. Later, these scientist—physicians stood trial in Nuremburg, Germany and were found guilty of war crimes. They were either executed or received long prison sentences. The other result of these trials was a code of ethics for medical experimentation with human participants. It is called the Nuremberg Code, and was adopted as a guideline for future research.

Ethics is the study of proper action that examines relationships between human beings and provides principles regarding how we should treat one another. The ultimate decision in ethical questions resides in judgments of value.

The Pursuit of Knowledge and Avoiding Harm

Ethical considerations of psychological experimentation have at their heart the idea that people participating in research should not be harmed. Of course, no psychological researcher seeks to harm a participant. However, there can be unexpected outcomes. Specifically, at the end of an experiment, participants should not be affected in a way that would result in a lower level of human functioning. This includes emotional distress. Ethical considerations must also look at the other side of the participant—scientist relationship. This includes the scientist’s right to know and to seek answers to questions. It would also be considered unethical to prevent a scientist from seeking knowledge without considering his or her rights. Thus, we begin with the rights of the scientist to know and to pursue knowledge and the rights of the participants to be protected from undue harm.

In most cases the scientist has a question that he or she wants to ask and that the participant is willing to help answer. Scientists have the right to study human functioning and answer appropriate questions. Participants have the right to be protected. In some cases, the participants learn something about themselves or about psychology from the experience, and they are glad to have participated. In brain-imaging studies, for example, participants often report that they enjoy seeing their brain activity (for example, functional magnetic resonance imaging (fMRI), electroencephalogram (EEG)) displayed. Other participants, like some travelers to a foreign country, enter the world of experimentation and leave it again without ever realizing it. Although ignorant of the underlying structure, they still leave with the experience of the event and may be changed by it.

If these experiences were always pleasant and any changes in the participant always positive, people would gladly participate in experiments, and scientists would face few ethical questions. However, at times the scientist may want to answer a question that requires that the participant experience psychological or physiological discomfort. These situations raise a number of questions:

1. What are the responsibilities of the scientist toward the participant?

2. What are the rights of the participant?

3. Are there guidelines for reconciling conflicts between the rights of the participant to pursue happiness and the rights of the scientist to pursue knowledge?

4. What type of relationship or dialogue would be most productive for helping the scientist and participant fulfill their needs and desires?

Since the 1950s, the American Psychological Association (APA) has published a set of guidelines. This is available online ( The federal government has adopted similar sets of guidelines for human and animal research, including the manner in which these studies should be reviewed.

Voluntary Participation and Informed Consent

What was unethical about the experiments at the Nazi concentration camps? It was not that human beings were given a virus, as almost all of our current procedures of preventive medicine (the polio vaccine is a historical example) require that the procedure eventually be tested on human beings. These physicians were convicted of conducting experiments without the consent of their participants. One of the first principles of research is that the participants must consent to being part of an experiment. Furthermore, they must also be informed of the experiment’s purpose and its potential risks. Thus, major ingredients in the dialogue between the scientist and the research participant are voluntary participation and informed consent.

Although a number of studies worldwide have ignored basic participant rights, the US Public Health Service began one particularly troubling study in 1932 that was originally designed to understand the long-term effects of syphilis. The participants were African-American men in Macon County, Alabama, who had syphilis. The people involved in the experiment were not told that they had syphilis and were denied any form of treatment. Although treatments for syphilis in the 1930s were not very effective, this study continued even after effective treatments were developed in the late 1940s. In fact, the study was not terminated until its existence was publicly exposed in a Washington newspaper in 1972, some 40 years later. This study not only violated basic ethical principles such as informed consent, but also put others, such as the participants’ families, at risk. It came to be called the Tuskegee experiment (

In the initial dialogue between the scientist and the prospective participant, the scientist must ask the participant to be a part of the experiment. This is the principle of voluntary participation. In essence, the voluntary participation principle suggests that a person should participate in an experiment only by free choice. In addition, this principle states that a participant should be free to leave an experiment at any time, whether or not the experiment has been completed.

As you think about voluntary participation, you will become entangled in the question of whether anyone can ever make a free decision and, if so, under what circumstances. As you might have realized already, this question becomes even more complicated for someone interested in developmental psychology, which requires research with children, or for someone interested in psychopathology, which requires research with patients who are mentally impaired. In terms of ethical concerns involving research with children, a number of recommendations have been put forward by the National Academies of Science, including how to obtain informed consent and ensure voluntary participation in research (Field & Behrman, 2004).

Assuming for a moment that someone can agree freely to participate in research, the scientist in the initial dialogue should inform the prospective participant about what will be required of him or her during the study. The scientist must also inform the prospective participant about any potential harm that may come from participation. Thus, the prospective participant must be given complete information on which to base a decision. This is the principle of informed consent. Consider the Tuskegee experiment—the principle of informed consent raises the issue of how much information about an experiment is enough.

From the principles of voluntary participation and informed consent, we can see that it is the initial task of the scientist to fully discuss the experimental procedure with prospective participants and to remind them that they are human beings who do not give away their rights just because they are taking part in a psychological experiment.

Confidentiality and Anonymity

In our society, research participants have the same rights during an experiment that they have outside the experimental situation, including the right to privacy. Most of us may initially consider the right to privacy as the right to spend time by oneself or with others of one’s choosing, without being disturbed. This is the external manifestation of the right to privacy. But there is also an internal or intrapersonal manifestation of this right (see Raebhausen & Brim, 1967). This is the right to have private thoughts or, as it is sometimes called, a private personality. This means that the thoughts and feelings of a participant should not be made public without the participant’s consent. It also means that a conversation between a participant and a scientist should be considered a private event, not a public one.

If that is the case, how can the scientist ever report her or his findings? There are two considerations that are part of the scientist’s responsibility to the participant: confidentiality and anonymity.

Confidentiality requires that the scientist not release data of a personal nature to other scientists or groups without the participant’s consent. Even during the experiment, researchers keep any personal data in a secure location and often destroy personal information once the experiment is completed.

Anonymity requires that the personal identity of a given participant be kept separate from his or her data. The easiest way to accomplish this is to avoid requesting names in the first place; however, there are times when this may be impossible. Another alternative is to use code numbers that protect the identities of the participants and to destroy the list of participants’ names once the data analysis has been completed.

The Ethical Relationship

Let’s remind ourselves that ethical questions have at their base issues of relationships and traditions. As scientists we ask what is and what ought to be our relationship with our participants and our society with regard to research. To answer this question, we stress that part of our ethical responsibility is to consult with others about our research. In this context, we would include the manner in which internal review committees evaluate the ethics of research and the guidelines (for example, those of the American Psychological Association and the federal government) used to direct our evaluations. With both humans and animals, we are only beginning to develop methods for the study of inner experience that can help to inform our ethical considerations. Part of our ethical responsibility is to use treatment approaches that have been shown to be effective, as described in the box: Applying Psychological Science—Empirically Based Treatments.

Applying Psychological Science—Empirically Based Treatments

Before the middle of the 20th century, very little formal research was performed to determine the effectiveness of psychological or medical interventions. Beginning in the 1950s and 1960s, there was a movement to determine the effectiveness of both in a scientific manner. In medicine, this came to be known as evidence-based medicine. This was the beginning of an approach that sought to replace opinion and tradition with solid scientific evidence. In psychology, the terms empirically based treatments or empirically based principles refer to psychological treatments and their aspects for which there is scientific evidence that the treatment is effective.

Three websites have been developed that list treatments for specific mental disorders.

✵ The first is maintained by the Clinical Psychology section of the American Psychological Association (APA) (

✵ The second is maintained by the US Substance Abuse and Mental Health Services Administration (SAMHSA), which is part of the US Department of Health and Human Services. This website contains a searchable online registry of mental health and substance abuse interventions that have been evaluated (

✵ The third website is devoted to effective treatments for children and adolescents ( This site is maintained by the Society of Clinical Child and Adolescent Psychology section of the APA.

There are a number of different treatments available when you visit a health care professional. You want to know what will work for you. That is the purpose of effectiveness research. Not all treatments work the same with every person. Thus, clinical researchers test treatments with a diversity of individuals. Such factors as gender, age, socioeconomic level, ethnic diversity, and so forth are important to consider.

As a society we want to focus more on approaches and principles for which there is scientific evidence. In determining what works there has been a willingness to integrate techniques from different theoretical approaches to treatment. The last chapter of this book will examine treatment for psychological disorders in greater detail.

Thought Question: We all act as scientists in our everyday lives. What situations in your life, for example, doing well in your studies or increasing your fitness level, could you tackle with effectiveness research? What series of experiments would you design to come up with empirically based principles for yourself? What would be the independent and dependent variables?

The Care and Use of Laboratory Animals

The issue of animal research has raised a number of questions over the past couple of decades. In fact, the American population is increasingly concerned about how animals are treated. In a 2015 Gallup poll, some 62% of Americans said that animals need protection from harm and exploitation, but it is appropriate to use them for the benefits of humans ( Some 32% of Americans took a stronger position and suggested that animals should receive the same protections as humans in terms of harm and exploitation. In specific situations, there was most concern for animals in the circus, sports events, and research. There was less concern for animals used as household pets or raised for food.

To protect animals from undue suffering or harm, a number of professional organizations have issued guidelines concerning the care and use of laboratory animals. The federal guidelines were issued by the US Public Health Service (PHS) in 1985 and updated in 1996 and 2002. In this policy statement, the PHS describes the manner in which animals are to be used and cared for.

It is clear that there are a number of emotional and polarized positions about the use of animals in research. The February 1997 issue of Scientific American offered a series of articles relating to animal research that illustrates well the opposing positions. For example, one position suggests that information learned from animal research is often redundant and unnecessary and may be misleading (Barnard & Kaufman, 1997). On the other hand, other researchers suggest that at least since the time of Pasteur, animal research has offered important breakthroughs in the treatment of disorders for both humans and animals (Botting & Morrison, 1997).

Those who support using animals in research point out that animal research is crucial to meet the goals of society. One propaganda poster points out that animal research has added 20 years (through medical research) to the lives of the people who protest against it. Of course, there is no one right answer for all situations. There are situations in which research with animals may not benefit society. As noted previously, even our research with humans has at times been inappropriate.

Today, scientists are seeking alternatives to using animals in laboratory research. One of the alternatives to laboratory research with animals is to study them in their natural environment. Other alternatives include computer simulations or the use of tissue cultures in biomedical research. In this book, you will notice that much of the animal research presented comes from an earlier period in the history of psychology. Technology now allows us to study human processes in ways not possible previously.

Humans and animals have a long-standing bond. Historical records suggest that animals have been kept as pets and used for herding sheep, plowing fields, and transportation for thousands of years. To protect their welfare, various organizations have been formed over the years, including the Society for the Prevention of Cruelty to Animals (SPCA), formed in England in 1824 and in the United States in 1866. Surprisingly, child abuse cases in the 19th century were initially brought to court under the laws against cruelty to animals because there were no similar laws for children!

Deception Studies

Deception research is any study in which the participant is deceived about the true purpose of the experiment or the experimental procedures. For example, if an experimenter wanted to examine the effects of anxiety on performance in college students, he or she might create an anxiety-provoking situation followed by a performance measure. One method used to create anxiety in college students is to administer a so-called IQ test that cannot be completed in the allotted time and to state that most college students who later go to graduate school have no trouble finishing this test. In the chapter on social process, you will be introduced to some classic studies in psychology that have used deception.

Another type of deception often used in medical research is the placebo treatment. For example, participants are given a pill made up of sugar or other inactive ingredients and told that it will cause certain physiological changes, such as reducing the number of headaches they have been experiencing. There is a long history that shows changes with placebos that cannot be explained by the medication itself (see Benedetti, 2009 for an overview). Thus, to give a medication with only inert ingredients could be considered deceptive. For this reason, most placebo studies tell the participants that they will either receive a placebo or a treatment with an active ingredient.

Most psychological studies that use deception try to create a situation in which the participant sees the world and what he or she is being asked to do in a certain way. Because it is impossible to obtain informed consent in deception research without compromising the study, is it unethical to perform this research? This is not an easy question to answer, and the professionals in the field currently fall into two camps.

The first group suggests that any deception is unethical, and thus no deception research is possible. The second group argues that certain types of deception research are necessary. This group further argues that, given appropriate safeguards, deception research is the only way to answer certain questions.

Once the experiment is concluded, however, it would be unethical to allow a participant to leave the experimental situation without a true understanding of what had taken place. The process of explaining the true purpose of the experiment afterward is called debriefing. In the debriefing procedure, the experimenter removes any misconceptions and offers a full discussion of the experiment. It is the goal of this dialogue to ensure that participants leave the laboratory with at least as much self-esteem and as little anxiety as when they came to the experiment (Kelman, 1968).

The debriefing process consists of two major aspects. First, the debriefing is an opportunity for the participants to tell the experimenter how they felt about being part of the experiment. This not only is good feedback for the experimenter but also allows the participants to express any self-doubt about their performance and deal with thoughts and feelings that arose during the experiment. Second, the debriefing is an opportunity for the experimenter to explain the study to the participants in greater detail.


✵ “Ethical considerations of psychological experimentation have at their heart the idea that people participating in research should not be harmed.” What four questions does every scientist need to consider in designing a research study?

✵ What are “voluntary participation” and “informed consent” in the context of scientific research? What are some of the specific issues they raise in terms of psychological research?

✵ How do confidentiality and anonymity figure into the experimenter’s responsibility to protect the research participant’s right to privacy?

✵ Why does the issue of animal research raise questions with the public? What are five reasons for conducting research with animals in biological psychology?


Learning Objective 1: Discuss the general process called science.

In general, there is no single scientific method, yet there is a general process called science. This process consists of experiencing the world and then drawing general conclusions (called facts) from observations. Sometimes these conclusions or facts are descriptive and can be represented by numbers. For example, we say that the moon is 238,000 miles from Earth or that the brain uses 20% of the energy produced by the body or that the average human heart rate is 72 beats per minute unless you exercise a lot and then it is probably about 50 beats per minute. Other times, these facts are more general and can describe a relationship or a process. For example, we say that it is more difficult to learn a second language after puberty than before or that as we age we hear fewer high-frequency sounds. Whatever the topic, the known facts about a particular subject are called scientific knowledge (see Ray, 2022 for more information).

Learning Objective 2: List the steps in the scientific method.

There are four stages to the scientific method: (1) develop an idea or expectation (hypothesis), (2) turn this hypothesis into an experiment, (3) evaluate the ideas and expectations about the world through observation and experimentation, and (4) draw conclusions or inferences about the ideas and expectations and consider the impact of the new information on theoretical conceptualizations. This is very different from the phenomenon of pseudoscience—or “fake” science—which presents information as if it is based on science when it is not. There are many research designs and which one to select begins with the question the scientist wants to answer. Some of the research designs used to study psychology include case study, naturalistic observation, correlational approaches, experimental method, and behavioral genetics designs.

Learning Objective 3: Explain how logic and inference are used to reach a conclusion.

Logic can help us answer questions of inference, which is the process by which we look at the evidence available to us and then use our powers of reasoning to reach a conclusion. Logical procedures are also important for helping us understand the accuracy or validity of our ideas and research. Using measures of two types of validity (internal and external), we logically design our research to rule out as many alternative interpretations of our findings as possible and to have any new facts be applicable to as wide a variety of other situations as possible. There is no “one perfect study”; designing and conducting research is always a trade-off between internal and external validity. Behavior can be described or measured in many ways; likewise, there are a variety of ways to measure a particular process. A useful way to begin analyzing the results of any experiment is to convert numerical data to pictorial form and then simply look at them. A second step is to calculate descriptive statistics for the sample—measures of central tendency and variability. The measure of central tendency that you use in your analysis—mean, median, or mode—depends on the question that you are asking. Measures of variability indicate how spread out the scores are. Measures of variability (or dispersion) include range, variance, and standard deviation. One characteristic of human beings is that we seek to determine what will happen next. If participants’ expectations (demand characteristics) or researchers’ expectations (experimenter effects) interfere with the influence of the independent variable, then the study could give inaccurate results. A related phenomenon is the placebo effect. To control for the placebo effect in research, various procedures have been used, the most powerful of which is to design a double-blind experiment.

Learning Objective 4: Describe the steps involved in designing an experimental study.

There are four steps to the experimental process that reflect the evolutionary nature of science: (1) the development of the hypothesis, (2) the translation of this hypothesis into a research design, (3) the running of the experiment, and (4) the interpretation of the results. Researchers take the results and interpretations of their studies and create new research studies that refine the previous hypotheses, and the cycle begins anew. One goal of experimental research is to determine the relationship between the independent variable (IV) and the dependent variable (DV). The less bias in terms of demand characteristics related to both the participant and the experimenter aids in creating a logical relationship between the IV and DV. Additional factors critical to sound inference are participant selection and assignment, the design of the experiment, and the interpretation of the relationship of the IV to the DV. The experimenter considers three hypotheses in interpreting whether the DV is related to the IV: null hypothesis, confound hypothesis, and research hypothesis. Scientists pay particular attention to research that has been evaluated by other scientists before it is published in a process called peer review, and journals that follow this procedure are called peer-reviewed journals. Replication of studies in different locations with different participants increases the certainty that the results found reflect the true nature of what is being studied.

Learning Objective 5: Discuss the ethical guidelines that psychologists follow in order to protect the rights of people participating in research projects.

Ethical considerations of psychological experimentation have at their heart the idea that people participating in research should not be harmed. In addition, research participants have a right to privacy including the right to a private personality. To protect those rights, participants must be informed of the experiment’s purpose and its potential risks (informed consent) and then voluntarily agree to participate in the experiment (voluntary participation). Confidentiality and anonymity are two additional considerations that are part of the scientist’s responsibility to the participant. Deception research and animal research are two areas of study in psychology in which specific ethical concerns have been raised. Guidelines for reconciling all conflicts between the rights of the participant to pursue happiness and the rights of the scientist to pursue knowledge are provided by such resources as the APA, and the federal government.

Study Resources

Review Questions

1. What does the author mean by “science is a combination of interaction with the world and logic”? What key role does doubt play in the process of science?

2. Design a series of research studies around a single question that uses each of the methods of science covered in this chapter: naturalistic observation, correlational approach, and experimental method.

3. Why can’t you design “the perfect study”? What trade-offs do you need to consider in designing an experimental study in the real world? What can you do to improve the quality of your study?

4. What roles do the following hypotheses play in interpreting experimental results:

a. null hypothesis?

b. confound hypothesis?

c. research hypothesis?

5. How is a scientist conducting psychology research like a detective solving a mystery? How are they different?

6. If we think about psychology research as an ethical problem, what are the rights of the research participant, and what are the responsibilities of the experimenter in ensuring the protection of those rights? What legal and ethical resources are available to guide this effort?

For Further Reading

✵ Kuhn, T. (1962). The Structure of Scientific Revolutions. Chicago, IL: University of Chicago Press.

✵ Mook, D. (2004). Classic Experiments in Psychology. Westport, CT: Greenwood Press.

✵ Ray, W. (2022). Research Methods for Psychological Science. Thousand Oaks, CA: Sage.

Web Resources

✵ APA ethics—

✵ Tuskegee experiment—

✵ Empirical therapy APA—

✵ SAMSHA approaches—

✵ Treatment for children—

✵ Gallop poll—

Key Terms and Concepts


blind controls

case study



confound hypothesis

confounding variables

control group

correlation coefficient

correlational research


deception research

demand characteristics

dependent variable (DV)

descriptive statistics

double-blind experiment




experimental design

experimental group

experimental method

experimenter effects

external validity





independent variable (IV)


inferential statistics

informed consent

internal validity



measures of central tendency




naturalistic observation

negative correlation

null hypothesis

operational definition


paradigm shift

peer review

placebo effect


positive correlation

private personality



random sampling




research hypothesis

right to privacy



scientific knowledge

scientific method

standard deviation





voluntary participation