Abstract
Adventures in Social Research (9th ed.) is a broad, but thorough, introduction to data analysis using the IBM SPSS Statistics software. Earl Babbie and his colleagues provide a thoughtful introduction for those beginning their scholarly endeavors while giving more advanced learners additional insight and guidance on future research. This is a good manual for undergraduate students. Using a step-by-step format, readers are shown how to open and save data, recode variables, and conduct quantitative analyses while giving a strong foundation of the research process in the social sciences. This edition features minor updates, including updating screenshots and instructions for the most recent edition of the SPSS Statistics software, using data from the 2012 General Social Survey (GSS), and separating a chapter examining significance tests into two chapters.
The book is structured into five parts, including (1) “Preparing for Data Analysis,” (2) “Univariate Analysis,” (3) “Bivariate Analysis,” (4) “Multivariate Analysis,” and (5) “The Adventure Continues.” Each part or section includes a number of individual chapters (ranging from two to nine chapters), with each successive chapter and section building on previously discussed content. For example, readers are shown how to recode and collapse variables (chapter 7) before they are shown how to create composite variables (chapter 8) as these tend to be more difficult for students to master. The authors introduce a concept or technique and provide a corresponding example with detailed instructions using data from the GSS, which is collected by the National Opinion Research Center at the University of Chicago and is available for free to students and educators online (http://gss.norc.org). The GSS is a nationally representative survey of American adults and includes a wide assortment of attitudinal and sociodemographic questions collected since 1972 (Smith et al. 2015). As the full GSS data set can be cumbersome, and inaccessible for those with student editions of the SPSS Statistics software, the authors provide subsets of the GSS data set online. Note, the web address in the manual is not current. Readers can access these data sets at https://study.sagepub.com/babbie9e. Occasionally, supplemental readings are also available online. For example, in their examination of how gender and age are associated with indicators of religiosity (chapter 10), the authors use the additional reading by Glock, Ringer, and Babbie ([1967] 2015) to provide a theoretical foundation for their hypotheses—particularly the deprivation theory of religiosity.
One the biggest strengths of this book is that readers are exposed to a variety of interesting and engaging examples, which should help students as they formulate their own research questions, including attitudes toward abortion, political orientation, and religiosity. For example, chapter 12 examines how gender, age, religious and political affiliation, and attitudes about sex are associated with views of abortion. For each covariate, the authors discuss a hypothesis, instruct the reader on how to examine the association with abortion attitudes in SPSS, outline their findings, and consider whether their hypothesis was confirmed or rejected. The authors provide helpful hints that make using the SPSS Statistics platform, and the research process overall, more approachable. For example, the authors suggest sorting variables alphabetically—which is not the platform default—to make navigating the data set easier. Throughout the text, the authors also provide writing examples of how to write up data sections, how to describe variables, and how to describe results from tables. These are helpful to students as they are learning how to discuss research findings. Each chapter concludes with main points, key terms, statistical commands, and helpful review questions.
Useful as out-of-class assignments, students should find review questions helpful for mastery of not only chapter content and techniques but also understanding some of the theoretical underpinnings of these techniques. For example, chapter 12 examines attitudes toward abortion. Speaking to differences in acceptance of abortion across political ideologies, the authors ask readers, “We found that liberals are more likely than conservatives to support abortion. Does this mean that Democrats are more likely to support abortion than are Republicans?” (p. 202). This question highlights that, although political affiliation and political ideology are similar, they are distinct constructs.
In addition to these questions, the authors include lab exercises that help the reader gain mastery of chapter techniques using a parallel research question or topic from the chapter. Lab exercises are sequential, allowing students to easily troubleshoot any difficulties by locating the analogous step in the example used in the chapter. For example, chapter 7 outlines five steps to recoding variables: preparing to recode a variable, recoding, checking the recode in the data editor, defining the variable, and running a frequency distribution (p. 107). In the chapter, the authors use church attendance, among others, as an example to teach readers to use these five steps. Using newspaper readership, as well as two other examples, the authors ask specific questions to reinforce these five steps. Instructors may find these exercises useful as in-class assignments or quizzes as the pages are easily detachable with each page perforated and three-hole punched. Yet, instructors may favor having students provide answers separately as some students may find the space provided insufficient. Instructors could also use the lab exercises provided as a template for their own class assignments or activities a number of ways. For example, instructors may use data from an alternative data set as the basis of the assignment (e.g., the public release of the National Longitudinal Study of Adolescent to Adult Health) or rely on different years or questions from the GSS.
In the first part, “Preparing for Data Analysis,” readers are provided with a succinct introduction to the research process. The authors briefly discuss concepts and theories and different types of variables. If Adventures in Social Research is used in an undergraduate research methods course, instructors may want to pair it with a traditional social research textbook—for example, The Practice of Social Research by Earl Babbie. In more advanced courses, such as a senior capstone course or an independent research project, this may not be needed. “Preparing for Data Analysis” also introduces the reader to the 2012 GSS, including subsets of the data set used throughout the book. Instructors using another data set as an example or in addition to the one presented in the book may want to introduce the data set concurrently with this section.
The second section, “Univariate Analysis,” outlines how to create new variables as well as how to articulate findings, through writing and graphically. Before any analyses, the authors provide the reader with an overview of the GSS and the basics of the SPSS platform, including opening and navigating data. Building on this, readers are shown how to run frequency distributions and a variety of descriptive statistics. Conveying analytic findings to a multitude of audiences is an important skill set for a number of careers and is paramount for those pursuing scholarly research further. The manual outlines a number of ways in which students can present their data graphically, including bar and line charts, using the SPSS Statistics chart editor. The authors provide a helpful chart that should aid students as they select an appropriate graphic for their variable. Students should find this helpful when constructing research briefs or research posters that may be a part of a research methods course or a discipline capstone course.
Recoding and creating composite measures can be difficult for students new to data analysis. The authors provide a number of helpful tips that should help students conceptualize collapsing variables and creating composite measures—for example, having students create a table with the values and labels for the original variable in one column and the values and labels of the new variable in another column (p. 107). Considering that students could be creating several new variables, the authors’ insistence on using labels when creating variables should help students navigate data more efficiently. Yet, I believe the authors are shortsighted in their guidance to students concerning data management. After making several changes to a subset of the GSS, named “DEMO.SAV,” in chapter 7, the authors appropriately instruct the reader to save the file. Although saying that the name of the new file name is at the discretion of the reader, the authors save their file as “DEMOPLUS.SAV.” Instructors should teach students to save data and output files using the project name and the date, at the least. For example, a project involving religiosity and political affiliation may be named as such: “RELIGPOLITIC_DATA_04082018.SAV.” That aside, the authors do a good job outlining how to recode and construct variables as well as how to apply the methods and techniques discussed to other topics that students should find interesting, including sexual permissiveness, prejudice, and raising children.
“Bivariate Analyses,” the third section, is primarily devoted to cross-tabulations but also covers correlation and regression analyses as well as tests of significance. Chapters 10, 11, and 12 examine variations in religiosity, political orientation, and attitudes toward abortion across several groups. Students should find the examples helpful when mastering how to create hypotheses and understanding the importance of distinguishing independent and dependent variables. A number of helpful suggestions and writing examples make understanding the results easier. For example, the authors suggest a 10 percent rule of thumb when determining when associations between two variables may be significant (p. 165). Some instructors may find the breadth of examples, although interesting, to be cumbersome for some courses. If faced with time constraints, instructors may find it beneficial to select chapter 10 for students to focus on, as the other chapters do not introduce new techniques. Students can easily refer to other chapters if they need additional practice. Subsequent chapters are devoted to examining the significance of these relationships across different types of variables. Depending on the scope of the course, some instructors may opt to focus on a specific set of techniques, such as Pearson’s correlations or chi-square. The authors introduce the reader to multiple regressions using ordinary least squares (OLS) regression models. Although briefly, the authors touch on the importance of writing out model equations; doing so is particularly important for those new to this analytic technique. The techniques discussed throughout this section would be easily paired with resources available in the American Sociological Association’s Teaching Resources and Innovations Library for Sociology (TRAILS). For example, using data from the 2016 GSS, Torres (2017) asks students to work in groups in order to evaluate information presented in online news sources. Selecting their own news sources and variables, students perform appropriate analyses to verify media claims.
The fourth section, “Multivariate Analysis,” builds on bivariate analyses from earlier chapters. Readers are first introduced to cross-tabulations with a control variable. The authors ask readers to consider alternative independent variables that might be related to some of the topics discussed in previous chapters. With the introduction of an additional variable, the authors draw attention to a common problem with data analyses: small cell sizes (p. 301). A possible solution, the authors hint, is to collapse variable categories. Although relevant for multivariate analyses, it would be helpful for students to find remedies to this problem earlier—perhaps when introducing cross-tabulations. Drawing on content discussed earlier, the authors introduce OLS regression analyses with controls briefly. Instructors should elaborate on this and, perhaps, introduce an activity that has students estimate group differences in predicted means.
In the last section, “The Adventure Continues,” the authors discuss possible next steps in the research process. This section is particularly helpful for more advanced students. Through an independent study or within a specific course, undergraduate students may engage in primary research. Briskly covering the steps of primary research—including designing a survey, data collection, entering data in SPSS Statistics, and analyzing data—the authors fail to discuss the institutional review board (IRB) process. The IRB process is often daunting for first-time researchers, and guidance is warranted. Some universities may require review even when students and researchers utilize secondary data. Despite this, the authors provide sample questions in an appendix that may be useful for those interested in a primary quantitative research project.
In sum, I recommend this book. Aside from the concerns I have discussed, both students and instructors should find this book to be a valuable resource. With the step-by-step format, wide assortment of relevant examples, useful tips, and lab exercises, this book should be a strong addition to an undergraduate research methods course as well as a helpful resource for more advanced courses, such as a capstone course or an independent study.