Principles of Research in Behavioral Science

Fourth Edition

Mary Kite, Bernard E Whitley

• Editore: Routledge
• Anno: 2018
• ISBN:
  9781315450063
• Edizione: 4
• Argomento: Psicologia
• Categoria: Ricerche e metodologie nella psicologia

This book provides a comprehensive overview of research methods in the behavioral sciences, focusing primarily on the conceptual issues inherent in conducting research. It covers topics that are often omitted from other texts, including measurement issues, correlational research, qualitative research, and integrative literature reviews. The book also includes discussions of diversity issues as they related to behavioral science research. New to this edition are chapter boxes that focus on applied issues related to each chapter topic. Throughout the book, readable examples and informative tables and figures are provided. The authors also take a contemporary approach to topics such as research ethics, replication research, and data collection (including internet research).

Part 1

Introduction

1

Behavioral Science

Chapter Outline

• Science
  • Goals of Science
    • Description
    • Understanding
    • Prediction
    • Control
  • Key Values of Science
    • Empiricism
    • Skepticism
    • Tentativeness
    • Publicness
  • Scientific Approaches to Knowledge
• Theories
  • Components of Theories
    • Assumptions
    • Definitions
    • Variables
    • Hypothetical Constructs
    • Propositions
  • Characteristics of Theories
    • Specification
    • Scope
  • Purposes of Theories
    • Organizing Knowledge
    • Extending Knowledge
    • Guiding Action
  • Criteria for Evaluating Theories
• Research
  • The Research Process
  • Evaluating Research
  • Inference in Research
• Theory, Research, and Application
  • The Interdependence of Theory, Research, and Application
  • The Uses of Behavioral Science Theory and Research
• Chapter Summary
• Suggestions for Further Reading
• Key Terms
• References

Why do people do what they do? How do biology, personality, personal history, growth and development, other people, and the social environment affect the ways people think, feel, and behave? As you have learned from your other behavioral science courses, there are many ways to go about answering questions such as these, ranging from the casual observation of daily life that leads to the development of personal theories of behavior (e.g., Wegner & Vallacher, 1977) to systematic empirical observation leading to the development of formal theories of behavior. The casual observation approach recognizes that we are all social scientists. As a result, throughout our daily lives, we use our “expectations, hunches, and hypotheses” (Maruyama & Ryan, 2014, p. 14) to make predictions about how other people will behave in various situations. Sometimes our personal ideas about the social world mirror the findings from social science research and sometimes they do not (Lilienfeld, Lynn, Ruscio, & Beyerstein, 2010). This imperfect match between personal beliefs and scientific knowledge exists because, as the philosopher Bertram Russell (1945) noted, scientists focus less on “what they believe” and more on “how and why they believe it” (p. 527). This “how and why” is reflected in scientific research methods—the set of rigorous tools and procedures that scientists use to answer research questions (Jaccard & Jaccoby, 2010). Another reason for the inconsistency between personal beliefs and scientific knowledge is that casual observation is prone to errors that stem from basic cognitive processes (e.g., Kahneman, 2011). Students of the scientific method are trained to understand how those errors influence judgment and learn how to critically examine their informal hypotheses to account for those biases (Maruyama & Ryan, 2014).

Behavioral science has three interconnected aspects: research that generates knowledge, theory that organizes knowledge, and application that puts knowledge to use. Most scientists have a greater interest in one of these aspects than in the others (Mitroff & Kilmann, 1978). Nonetheless, one’s complete development as a scientist requires an understanding of all three (Belar & Perry, 1992). As the title of this book indicates, we will explore the research aspect of behavioral science in detail. But before doing so, we will put research into context by reviewing the nature of science and theory and by examining how research, theory, and application interrelate.

Science

Science is a systematic process for generating knowledge about the world. Science has three important aspects: goals to be achieved, key values to be enacted, and perspectives on the best way to go about generating knowledge. We begin by reviewing each of these aspects.

Goals of Science

Behavioral science has four goals: the description, understanding, prediction, and control of behavior.

Description. As a goal of science, description has four purposes. The first is to define the phenomena to be studied. If you were interested in studying memory, for example, you would need to start by defining memory and by describing what you mean by that term. Do you mean the ability to pick out something previously learned from a list, as in a multiple-choice test, or the ability to recall something with minimal prompting, as in a short-answer test? Or are you interested in retrospective memories, such as the experiences adults recall from their childhood? It is important to hone in on your definition because your research question and your approach to answering that question are linked to the type of memory that interests you. Thus, the second important purpose of description is clearly differentiating among closely related phenomena; by doing so, you can be certain you are studying exactly what you want to study. Environmental psychologists, for example, distinguish between population density, the number of people per square meter of space, and crowding, an unpleasant psychological state that can result from high population density (Stokols, 1972). High population density does not always lead to feelings of crowding. People who are feeling positive emotions rarely feel crowded even in very densely populated situations—think about the last time you were enjoying a large party in a small house. We return to this definitional aspect of the descriptive goal of science later, when we discuss the components of theories.

The third purpose of description as a goal of science is recording events that might be useful or interesting to study. Let’s assume, for example, a friend of yours has been in an automobile accident. There is a trial to determine liability, and you attend the trial to support your friend. As the trial progresses, something strikes you as interesting: Although four witnesses to the accident had equally good opportunities to see what happened, their memories of the event differ significantly. Two witnesses estimated your friend’s car was moving at 30 miles per hour (mph) when it hit the other car, whereas two others estimated the car’s speed to be 40 mph. This inconsistency piques your curiosity, so you attend other trials to see if such inconsistencies are common, and you find they are. You have now described a phenomenon: Eyewitnesses can produce very inconsistent descriptions of an event.

Finally, science works to describe the relationships among phenomena. Perhaps in your courtroom observations you noticed that some kinds of questions led to higher speed estimates than did other kinds of questions. This discovery of relationships among phenomena can help you understand why certain phenomena occur—the second goal of science.

Understanding. As a goal of science, understanding attempts to determine why a phenomenon occurs. For example, once you determine eyewitnesses can be inconsistent, you would want to know why the inconsistencies exist—that is, what causes them. To start answering this question, you might propose a set of hypotheses, or statements about possible causes for the inconsistencies, that you could then test to see which (if any) were correct. For example, given courtroom observations, you might hypothesize that the manner in which a lawyer phrases a question might affect a witness’s answer so that different ways of phrasing a question cause different, and therefore inconsistent, responses.

But how can you be sure a particular cause, such as question phrasing, has a particular effect, such as producing inconsistent answers? To deal with this question, the 19th-century philosopher John Stuart Mill developed three rules, or conditions, for causality; the more closely a test of a hypothesis meets these conditions, the more confident you can be that the hypothesized cause had the observed effect. The three rules are:

• Covariation: The hypothesized cause must be consistently related to, or correlated with, the effect.
• Time precedence of the cause: During the test of the hypothesis, the proposed cause must come before the effect.
• There must be no plausible alternative explanation for the effect: The hypothesized cause must be the only possible cause present during the test of the hypothesis.

Although the principle of time precedence may seem obvious—in our example, the lawyer’s question came before the witnesses estimated the speed at which the cars were moving—you will see in the next chapter that some research methods cannot establish with certainty that a hypothesized cause did, in fact, come before the effect. Ruling out plausible alternatives is the most complex of Mill’s conditions. For example, if you test your question-phrasing hypothesis by having one person word a question one way and another person word it another way, you have a problem. You have no way of knowing whether any response differences that you observe are caused by the different phrasings (your hypothesis) or by some difference between the two people asking the question. For example, perhaps a friendly questioner elicits one kind of response and an unfriendly questioner another. In contrast, if the same person asks the question both ways, differences in personality cannot affect the answer given. We address this problem of alternative explanations in more detail in Chapter 7.

Bear these rules of causality closely in mind; they will play an important role in the discussions of research strategies in the next chapter and throughout this book. By the way, if you tested the question-phrasing hypothesis according to these rules, as did Loftus and Palmer (1974), you would find that question phrasing does affect responses. Loftus and Palmer had people watch a film of a moving car running into a parked car. Some of the people estimated how fast they thought the moving car was going when it “smashed” into the other car; other people estimated how fast it was going when it “contacted” the other car. The average speed estimate in response to the “smashed” question was 40.8 mph; in contrast, the average estimate in response to the “contacted” question was 31.8 mph.

The scientific goal of understanding also looks for relationships among the answers to “why” questions. For example, you might want to know how well your question-phrasing hypothesis fits in with more general principles of memory (Penrod, Loftus, & Winkler, 1982). An important aspect of scientific understanding is the derivation of general principles from specific observations. Thus, if a thirsty rat is given water after pressing a bar, it will press the bar again; if a child is praised after putting toys back in the toy box, he will do so again; if a stand-up comedian tells a joke that gets laughs, she will keep that joke in her routine. These observations can be summarized with one general principle: Behavior that is rewarded will be repeated. Systems of such general principles are called theories, which are discussed later in this chapter. Once general principles are organized into theories, they can be applied to new situations.

Prediction. As a goal of science, prediction seeks to use our understanding of the causes of phenomena and the relationships among them to predict events. Scientific prediction takes two forms. One form is the forecasting of events. For example, the observed relationship between Graduate Record Examination (GRE) scores and academic performance during the first year of graduate school lets you predict, within a certain margin of error, students’ first-year grades from their GRE scores (Wendler & Bridgeman, 2014). When other variables known to be related to graduate school performance, such as undergraduate grade point average (GPA), are also considered, the margin of error is reduced (Kuncel, Hezlett, & Ones, 2001). Notice that such predictions are made “within a certain margin of error.” As you know from your own experience or that of friends, GRE scores and GPA are not perfect predictors of graduate school performance. Prediction in behavioral science deals best with the average outcome of a large number of people: “On the average,” the better people do on the GRE, the better they do in graduate school. Prediction becomes much less precise when dealing with individual cases. However, despite their lack of perfection, the GRE (and other standardized tests) and GPA predict graduate school performance better than the alternatives that have been tried (Swets & Bjork, 1990).

The second form taken by scientific prediction is the derivation of research hypotheses from theories. For example, people are very accurate in recognizing certain emotions, such as anger, happiness, and surprise, displayed on an isolated face (Ekman, 2016). Other emotions, such as pride and disgust, can be more difficult to categorize. Researchers have proposed that people rely on the situation in which the emotion occurs to interpret such facial expressions (Barrett, Mesquita, & Gendron, 2011). Take a moment to visualize the face of an athlete who has just won a challenging race. If you were shown such ...