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

to performance in graduate school. Therefore, there is controversy about what consideration, if any, GRE scores should have in admissions decisions. This controversy centers on:construct validity.criterion validity.test–retest reliability.internal reliability.

      6 Ranking a group of cities in terms of “quality of life” would be an example of measurement of a(n) _____________ scale of measurement.nominalordinalintervalratio

      Chapter 3 Describing Data With Frequency Distributions and Visual Displays

       After reading this chapter, you will be able to

       Differentiate between a frequency distribution table and a grouped frequency distribution table

       Discuss how bar graphs, histograms, and frequency polygons are used to communicate frequency distributions visually

       Interpret frequency distribution information conveyed in bar graphs, histograms, and frequency polygons

       Discuss how bar graphs, scatterplots, and line graphs are used to communicate relationships between variables visually

       Interpret relationships between variables that are displayed using bar graphs, scatterplots, and line graphs

       Construct visual displays of frequency data and relationships between variables using SPSS

      In the previous chapter, we learned about basic considerations in conducting quantitative research. With these basic concepts in hand, we can now begin discussing what to do once we have collected data from a sample. In this chapter, we will discuss a study by Laura Wendt (2013) that examined predictors of academic burnout in first-year and senior-year college students. We will continue to use this same study and its data in the next three chapters as we extend our discussions of descriptive statistical tools in those chapters. In this chapter, we will first discuss this study, the constructs it examined, and how the constructs were operationalized. In subsequent sections of this chapter, we will discuss ways that researchers can summarize large amounts of data, both with tables and with visual displays. After learning about three commonly used tools researchers use to visually display data and relationships between variables, we will conclude this chapter by using the software program Statistical Package for the Social Sciences (SPSS) to construct these commonly used tools. Be aware that you have access to Wendt’s (2013) data in the file titled “Wendt’s data.sav”. We will use these data later in this chapter and in the next chapter.

      The Study

      In her research, Wendt (2013) had 46 college first-years and 62 college seniors complete a packet of surveys. In this packet were measures of demographic variables, including participant sex, age, and year in college. In addition, the packet contained three additional measures. First, there was a 15-item measure of academic burnout (Schaufeli, Martinez, Marques Pinto, Salanova, & Bakker, 2002). By “burnout,” we mean prolonged stress that causes people to feel exhausted, cynical, and as if they cannot accomplish their responsibilities effectively. Burnout tends to result not from working too hard per se but from feeling that one does not reap the benefits associated with one’s efforts. For example, if you study a lot and make good grades, you’re not likely to burn out because despite working hard, you’re reaping the rewards of your hard work. The burnout questionnaire contained items such as “I have become less enthusiastic about my studies” and “Studying is really a strain for me.”

      Second, there was a 13-item measure of role overload (Reilly, 1982). By “role overload,” we mean the extent to which a person feels he or she can manage the demands with the different roles that must be played. For instance, in your role as a student, certain types of behaviors (e.g., attending class, studying, and making presentations) are expected. Likewise, in your role as a friend to other people, different types of behavior are expected (e.g., returning texts, talking, and providing advice as needed). Role overload occurs when people believe they are having a difficult time meeting all of these demands. This questionnaire contained items such as “There are times when I cannot meet everyone’s expectations” and “I can’t ever seem to get caught up.”

      Finally, there was an 11-item measure of dysfunctional perfectionism (Khawaja & Armstrong, 2005). Being a perfectionist can of course be a good thing as it often drives people to do well at work and in school. However, perfectionism becomes dysfunctional when it creates an obsession with avoiding failure and mistakes, setting unattainable goals, and being unable to learn from feedback. For instance, it is good to proofread a paper, but if you spend so much time proofreading that you don’t study for a test the next day, that’s dysfunctional perfectionism. This measure contained items such as “Even when I do something very carefully, I often feel that it is not quite right” and “I set higher standards than most people.”

      Students responded to the burnout, role overload, and perfectionism items using a 1 (strongly disagree) to 5 (strongly agree) response range. Thus, burnout scores could range from 15 to 75; role overload scores could range from 13 to 65; and dysfunctional perfectionism scores could range from 11 to 55.

      Before we begin using Wendt’s (2013) research to illustrate basic ideas in this chapter, let’s make sure we’re thinking the same way about why we are discussing these ideas. In this chapter and the two chapters that immediately follow it, we are going to discuss ideas about how to use descriptive statistics. Recall from Chapter 1 the notion of descriptive statistics. Specifically, we want to make sense of (i.e., describe) large amounts of data so that people can understand them easily. The purpose of frequency distributions, visual depictions, and SPSS is to help us make sense of large amounts of data. I think of the tools we are about to discuss as being to a psychologist what clay and a kiln are to a ceramicist. Without clay and a kiln, there can be no pottery. Without understanding descriptive statistics, there can be no quantitative psychological research.

      Frequency Distributions

      In Wendt’s (2013) research, there were a total of 108 respondents. Each respondent had a score on each variable in this study, which means that there were a lot of data that she needed to organize. In this section, we will focus on the scores from the measure of burnout. Table 3.1 contains the burnout score for each respondent. Recall that scores on the burnout measure could fall between 15 and 75. If you can meaningfully summarize Table 3.1, you’re more insightful than I am. What we need to do here is somehow organize these 108 burnout scores so that people can make sense of them. The purpose of a frequency distribution is to organize a large amount of data in a format that people can understand quickly. There are two formats that a frequency distribution can take: a frequency distribution table and a frequency distribution graph. Let’s use the 108 burnout scores from Wendt’s research to illustrate both formats.

      Frequency Distribution Tables

      Again, if you can make sense of the burnout scores in Table 3.1, I am happy for you. However, most of us need some tools to organize these data better than they are currently presented. We can do this in three ways by using tables of the scores. We will discuss each of these three types of tables in the order of increasing organization of our scores.

      First, let’s organize our 108 burnout scores from highest to lowest. We’ve done this in Table 3.2. Now we can now get a sense of what constitutes “high” and “low” burnout scores. For instance, although scores could go as high as 75, the highest score in this sample is 55. This means that there weren’t any “completely burned out” students