Interpreting and Using Statistics in Psychological Research. Andrew N. Christopher

context, they are not of interest to the researchers. Therefore, we want to control for their influence on how people in the sample behave, so that we can isolate the effect of the independent variable. Through the process of random assignment, we can minimize the influences of variables (e.g., where people grew up, when people grew up, and socioeconomic status) other than the independent variable. In doing so, any effects we find can be linked to the independent variable.

      Suppose we find that students who engaged in aerobic exercise throughout the semester had higher grades at the end of that semester. By using experimentation, which involves manipulating at least one variable, and by using random assignment of participants to groups created by that manipulation, we can draw cause-and-effect conclusions between behaviors. We will explore statistical tools that are often used with experimental data in Chapters 7 through 11.

      Explanatory research: draws cause-and-effect relationships between variables.

      Independent variable: variable that a researcher manipulates (changes) to create experimental groups (conditions). It should affect subsequent behavior or mental processes.

      Dependent variable: behavior that results from the independent variable.

      Random assignment: uses a random process, such as flipping a coin, to put members of a sample in one of the groups (conditions) in an experiment. Its purpose is to minimize preexisting differences among members of the sample so that researchers can be confident of the effects of the independent variable.

      Goal: To Apply

      Finally, applied research does not have specific research methods associated with it. Rather, it makes use of the findings from the methods described previously and uses them in specific contexts. For instance, when you listen to the weather forecast (a prediction) for the next day, you will apply that information by dressing accordingly. Likewise, we already know that the manner in which information is presented influences how people respond to that information (called the framing effect). The makers of Ruffles potato chips were clearly aware of the framing effect when they labeled the cooking method of “baked” on those bags of chips, but they did not label the cooking method of the “fried” chips.

      To continue our example of health-related behaviors in college students, many colleges and universities are keenly interested in promoting physical and psychological well-being in their student populations. Like all organizations, these schools face budgetary constraints. Therefore, they want to maximize the desired outcomes (well-being in their student populations) within those financial limitations. To do so, many schools rely on the sorts of research studies we have described in this part of the chapter. For instance, one college wanted to update its aerobic exercise machines. It first conducted a survey of its students and faculty by asking these people to complete a questionnaire. Based on the results of this survey, the college invested a portion of available funds and updated some of the aerobic exercise equipment. Then, after this new equipment was available for use, it conducted naturalistic observation to learn if in fact the equipment was being used as indicated it would be used on the questionnaire. When those naturalistic observations suggested heavy use of the new equipment, the college immediately invested the remaining portion of its funds for this purpose.

      Applied research: uses descriptive, predictive, and explanatory research methods to answer specific research questions in specific contexts.

      Learning Check

      1 What is the difference between naturalistic and laboratory observations?A: Naturalistic observations occur in the normal environment in which a behavior occurs. Laboratory observations occur in a more controlled setting. For instance, if we want to study driving behaviors, we could conduct naturalistic observation by watching and recoding driving behaviors on the interstate. We could conduct laboratory observation by using a driving simulator in a research lab.

      2 Case studies are not used as much as other descriptive research methods. Why is this the case?A: Case studies tend to be used to study rare phenomena. For instance, much of what we know about the effects of strokes we gained through case studies of people who have actually experienced a stroke. Depending on where precisely in the brain a stroke occurred, a case study allows us to learn what happens when that part of the brain is damaged.

      3 Explain the difference between a positive correlation and a negative correlation.A: A positive correlation exists when two variables tend to “move in the same direction.” A negative correlation exists when two variables tend to “move in opposite directions.” For example, there is a positive correlation between the number of hours that students study for tests and their scores on those tests. There is a negative correlation between the amount of alcohol consumed the night before a test and scores on that test.

      4 I do a study in which I compare students’ political beliefs at the start of their first year of college and again when they graduate from college. Explain why this study is a quasi-experiment.A: As year-in-school is a participant variable, I cannot randomly assign people to be first-years or seniors. These are naturally occurring groups to which people belong.

      5 How does random assignment allow an experimenter to claim that an independent variable influenced a dependent variable?A: Random assignment minimizes preexisting differences between people who are in an experiment. Therefore, with these differences minimized, researchers can isolate the effects of the independent variable on the dependent variable.

      6 A college registrar did a research project and found the typical 18- to 22-year-old college student is mentally more alert later in the morning than earlier in the morning. Therefore, her college decides not to schedule any classes before 11 a.m. How is this an example of applied research?A: The college is taking a research finding (that students aren’t as mentally alert early in the morning) and then using it to implement a policy in a specific situation.

      Statistical Thinking: Some Basic Concepts

      One commonalty in the flaws in our thinking is that they all, to varying extents, rely on our personal experiences to learn about the world. Personal experiences can sometimes be helpful. For instance, when I was three years old, I touched a hot stove burner. I learned right then and there never to do it again. However, just because our experiences may provide us with information, it does not mean that such information is necessarily correct or true of people in general. We use statistics not to learn about any one person but to learn about people (and animals, chemicals, and many other topics of study) in general. Let’s discuss, and in some cases, reinforce, some extremely important statistical concepts that will appear throughout this class.

      Parameters Versus Statistics

      We discussed the notions of population and sample previously in this chapter. We will expand on them here. Recall that a population is the entire group of people we want to learn about, and a sample is a subset of people drawn from that population that is intended to represent the characteristics of its population. In research, the purpose of the sample is to learn about characteristics of the population. As we said, when we say “characteristics of the population,” we are talking about variables, which again, are qualities that have different values or change among individuals.

      As we discussed in this chapter, in most walks of life, we must operate with information about a sample from our personal experiences, and that can be problematic when drawing conclusions about the world. Variables, typically expressed quantitatively, that describe a population are called parameters. An example of a parameter would be the odds of dying in a terrorist attack and the odds of dying from a fall at home. However, we rarely, if ever, know and use parameters in our daily thinking. Rather, we rely on samples. This can be a problem when our goal is to be objective. When we want to be objective, we need some help. Therefore, we make use of statistics, which are accepted quantitative procedures that allow us to organize, summarize, and interpret