Compassionate Statistics

Applied Quantitative Analysis for Social Services (With exercises and instructions in SPSS)

Vincent Faherty

• Publisher: SAGE Publications, Inc
• Year: 2007
• ISBN:
  9781483316949
• Edition: 1
• Topic: Social Sciences
• Subtopic: Social Work
• Index: Social Sciences

Compassionate Statistics: Applied Quantitative Analysis for Social Services (With Instructions for SPSS 14.0) is an attempt to "de-mythologize" a content area that is both essential for professional social service practitioners, yet dreaded by some of the most experienced among them. Using friendly, straightforward language as well as concrete illustrations and exercises from social service practice, author Vincent E. Faherty catapults students and experienced professionals to a pragmatic level where they can handle quantitative analysis for all their research and evaluation needs.

1

Introduction, Overview, and Nondefinitions

Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: “There are three kinds of lies: lies, damned lies, and statistics.”

—Mark Twain, Autobiography, 1904,
cited on www.quotegarden.com/statistics

Welcome to Compassionate Statistics: Applied Quantitative Analysis for Social Services (With exercises and instructions in SPSS), a book obviously about the one subject that is universally dreaded by most social service professionals—statistics. And rightly so. After all, no matter what your official title is—advocate, case manager, caseworker, counselor, gerontologist, probation officer, parole officer, social worker, or something else—you are in this high-pressure, tragically misunderstood, and severely underpaid profession precisely because you like people and not because you love to crunch numbers like your annoying relative, the accountant, who always seems to find you at family gatherings, just to harangue you with his latest insights into public welfare fraud and ply you with his ideas on how to fix it. So, you look at yourself, happily, as not just another number cruncher. Your goal is to provide the absolute best, state-of-the-art services to your clients who are struggling with horrific personal, economic, and social difficulties every day of their lives.

Then again, you might be a supervisor or upper-level administrator in one of those social service agencies that employs those social service professionals in the front lines of the vast array of community-based organizations. Your mission is distinct from that of your staff. What keeps you awake at night is how you can make available to those vulnerable clients a staff that is highly trained and sensitively responsive to their particular needs. And, in the midst of all these human resources issues, you need to run an organization that is effective and efficient in its current operations and yet open to change and dynamic development.

Or, perhaps, you are a baccalaureate or beginning graduate student in one of the thousands of academic programs in colleges and universities throughout society. Your need is different, again. In your busy schedule, where you juggle life and education, sometimes precariously, your goal is to finish your professional education so you will be adequately prepared to join your professional colleagues on those front lines of service. Amid all the exciting content you are exposed to during your years of study, it is entirely understandable that you might hesitate, ever so slightly, and wonder to yourself, as you register for those required courses in research methods and statistics, why your professors are making you take courses with such irrelevant nonsense. After all, you are in your program because you want to help people, not just do research and count numbers!

Well, if you picture yourself in any one of those scenarios just described, my message is simple and straightforward. With apologies to the creators of the Budweiser beer commercial—this book’s for you!

Compassionate Statistics: Applied Quantitative Analysis for Social Services has been designed to serve you, the professional social services student, practitioner, or administrator, as you pursue diligently your many roles and responsibilities in today’s society. This book will get right to the core of the essential issues you will need to deal with as you collect, process, and analyze those relevant sets of quantitative data in your work. This is not a watered-down version of statistics, but it is, admittedly, an attempt to go beyond the theoretical material usually found in classic statistical textbooks and lift you, very quickly, to the pragmatic, day-to-day realm, where you are right now. In many senses, this book is an extension of the often-stated value premise of the need to start where the client is. In this situation, then, Compassionate Statistics: Applied Quantitative Analysis for Social Services is starting where, hopefully, you are.

One final point: This introductory discussion is not meant to disparage in any way the wide array of excellent textbooks on statistics, especially those designed for the social service community. They are important contributions to the professional literature and should be consulted by you, as needed, whenever this book assumes too much or moves too quickly for you from the abstract to the concrete. You are encouraged to keep those standard texts in your library and use them, as appropriate, in conjunction with this book. It goes without saying, of course, that you should also hold very close to you any textbooks that you have used (or are using presently) that discuss the methods and processes of social science research and evaluation in general. Consider those as essential tools to conduct any empirical research or program evaluation project.

Structure of the Book and of Each Chapter

Beginning in the next chapter, Chapter 2, you can expect to find the following material in sequential order.

• Statistical Content. Against a background that includes a minimum of theory but a maximum of social services examples, the major statistical issues/themes will be presented. This is a fairly standard list of issues/themes that will provide you with the knowledge necessary to conduct basic, firstlevel research and evaluation projects that require some level of quantitative analysis. Chapters will be kept mercifully short and to the point in order to facilitate the process of engaging you in the acquisition of some new (and difficult) content or, perhaps, reinforcing content that you experienced years before. What you will not find in the content is any statistical formula, which would allow you to compute a statistical test by hand. This decision to keep Compassionate Statistics: Applied Quantitative Analysis for Social Services free of any mathematical equations or formulas may be criticized by some statisticians as being overly simplistic and, perhaps, even misleading. Hopefully, that is not true. This book on applied statistics is intended, as stated above, to be a supplementary text, or perhaps an introductory one, that offers you practical knowledge and functional skills that you can apply immediately in your role as a student or as a working social services professional. Teaching you to access computerized software in order to analyze the data, rather than have you process statistical formulas on calculators, was a decision made carefully and is based on the goal of helping you use statistics in your work, rather than only understand its potential.

• Summary Points to Remember. Each chapter content discussion will end with a brief, targeted summary section designed to help you recall the essential elements of each chapter.

• Case Illustrations. Using selected material drawn from the professional literature, four actual examples of the statistical issue or procedure discussed in the chapter will be included so you can envision how that particular statistical material did actually appear in a professional journal or book. Thus, you will encounter real data and see exactly how those data were analyzed and presented in print. Additional explanatory material is added after each case to clarify further the statistical procedure under discussion. The first of the four case illustrations is presented at the very beginning of each chapter (with the exception of Chapters 1 and 8) so as to provide you with a concrete example of what will be discussed in that particular chapter. Three additional case illustrations appear toward the end of each chapter, after the content has been presented.

• Activities. To immediately fortify your newly acquired knowledge with practice, you will be challenged to conduct a series of relevant statistical exercises with data provided in one or more of the data sets in the appendix section of the book. For some chapters, you may be asked to enter these raw data into a new Statistical Package for the Social Sciences^® (SPSS) file that you will save for future use; in other chapters, you may be directed to the Sage Web site, where you can download the data to your own computer for statistical analysis in SPSS. There also exists the possibility that your instructor may modify these activities, or add new ones, for your unique learning experiences.

• SPSS Procedures. To assist you in completing these activities, you will have access to clear, step-by-step SPSS procedures, including appropriate computer screen images, relating to the specific content of each chapter. In case you have never used SPSS or are a bit rusty on its operations, you will find a thorough, general introduction to SPSS in Appendix A: Getting Started With SPSS at the end of this book. Please note, however, that the SPSS procedures in each chapter, as well as in Appendix A, are not intended to replace a formal SPSS manual or to substitute for the extensive Help options provided within the online SPSS software program itself. The chapter-by-chapter SPSS procedures and screen shots will assist you in completing the assigned activities, but you might need additional SPSS material to analyze your individual research and program evaluation projects. While all SPSS computers screen shots for this book were created in SPSS 14.0, they should be useful for future editions of the SPSS software with only slight visual modifications.

• References. Finally, at the end of each chapter, you will find a list of citations, in the American Psychological Association (APA) style format, for each professional article used in that chapter. This will help you easily find the full text of those articles if you ever wish to read about the entire research project beyond the small excerpt included in each case illustration.

Some Useful Nondefinitions

The phrase nondefinitions may sound strange, and that is intentional. They are designed to introduce you to some highly complex, and at times confusing, quantitative concepts that are fundamental to your understanding of statistics. Nondefinitions help you ease into some pretty deep waters—sort of like when you get your feet wet before diving into that crashing wave at the beach. By the way, at the end of this book, there is Appendix F: Glossary of Terms, which includes all the technical terms used throughout the chapters. The glossary is the closest thing to a list of formal definitions you will come across. If you find yourself wanting more details about some term, you could return to your research methods book or to that dog-eared standard statistics book you might still have lying around someplace, or you could log on to the Internet (try google.com, ask.com, msn.com, or wikipedia.com).

Hopefully, what you will find helpful in addition to the glossary is a less technical, example-riddled, almost breezy discussion of these important terms and phrases, similar to having a simple conservation with a friend. That’s what a nondefinition is in this book. They are listed here alphabetically, not necessarily in the order of importance. The suggestion is that you read through these nondefinitions once now, as you start this book, and then remember that they are here in Chapter 1, for your review, as you travel through the rest of the chapters. Happy journey!

Degrees of Freedom

This is a classic example of a statistical word or phrase that does not mean what it sounds like it should mean. Freedom in this case refers to the number of ways a variable is able to differ or fluctuate. When you put the variable gender with two possible categories (female, male) in the same data set as the variable location also with two categories (urban, rural), you have several possible combinations (female urban, female rural, male urban, male rural). The degrees of freedom (noted as df) should always be reported when you note the output of a statistical test. Now, relax. The SPSS in most cases gives you the df number in the output, so all you have to do is note it and record it correctly. The only situation in which you have to do your own computation of the df is whenever you run a correlation test. Stay relaxed … it is easy. The df is the number of subjects minus two (N – 2 = df). Remember that and you will never make a mistake when you are computing either of the two correlation tests (Pearson’s r or Spearman’s rho).

Dependent and Independent Variables

Think of any typical cause/effect situation. For example, if your right elbow radiates a dull ache all day long whenever it is raining, you might logically conclude that the dampness in the air is somehow causing your elbow to experience such a pain. In any study of two variables, the one that is influencing, pushing, or exerting some power over the other variable is called the independent variable. The independent variable is usually noted as X in a figure or diagram. The dependent variable, then, is the variable being pushed or being influenced in some way. The dependent variable is what you measure as a researcher and is usually noted as Y in figures or diagrams. In the hypothesis, “Women in this group of applicants will score an average of 5 points higher on the state driving test than men,” the independent variable would be gender, and the dependent variable would be the test score on the state driving test.

Descriptive and Inferential Statistics

The words say exactly what these terms mean. You will cover a large number of statistical procedures (tests) in this book, some very simple and others quite complex. These can all be divided neatly into two categories or groups of procedures. Those procedures that merely summarize, condense, or describe one or two variables at a time are called descriptive statistics. You l...