Social Statistics

Managing Data, Conducting Analyses, Presenting Results

Thomas J. Linneman

• Publisher: Routledge
• Year: 2017
• Print ISBN: 9781138228634
• eBook ISBN: 9781315391809
• Edition: 3
• Topic: Social Sciences
• Subtopic: Social Science Research & Methodology
• Index: Social Sciences

With a clear and engaging writing style and strong examples from the real world, this text covers current statistical techniques at an introductory level and emphasizes the clear presentation of results to a variety of audiences, making the course more useful to students and their careers. Interconnection features among chapters help students understand how all of the techniques fit together. Using varied data sets, the text features a highly rated companion website that includes videos of the author offering step-by-step explanations of how to carry out the techniques, interpret the results, and present them to varied audiences.

NEW TO THIS EDITION

• More inter-chapter connections have been added to improve students' conceptual learning.
• Several examples (on immigration, health, and civil rights) now permeate the text for easy comparison of techniques across chapters.
• The section on managing data is considerably expanded to cover topics such as finding new sources of data, dealing with missing data, and how to combine data reliably.
• Very current examples from the scholarly literature from criminology, education, and health show how researchers use each chapter's techniques to tell compelling stories.
• Instructors can choose from a variety of greatly expanded materials to enhance their lectures: engaging animations of key concepts; dynamic demonstrations of how statistics change in line with the data; short lectures on difficult-to-explain topics; and in-class exercises that will help students learn how to make sense of statistical results.

Chapter 1
Life in a Data-Laden Age

Finding and Managing Datasets

This chapter covers ...

• . . . what data look like in their raw form within a dataset
• . . . how to work with data to get them ready to analyze
• . . . the wide variety of datasets that are readily available for analysis
• . . . how to build an additive index and check if it is acceptable
• . . . the newer forms that data take, from Internet databases to media analyses
• . . . types of variables used in statistical analysis
• . . . a classification of statistical procedures we’ll cover in this book
• . . . an example of how researchers used an Internet dating site to study multiracial people
• . . . an example of how researchers used Google and Twitter to study media effects.

Introduction

Though considered a rude question in other parts of the world, a very typical first question at American parties when people are meeting other people for the first time is “What do you do for a living?” I occasionally attend parties, and so I have had the following conversation many times:

OG (other guest): What do you do for a living?

TL (that’s me): I’m a professor.

OG: Really? What do you teach?

TL: Introductory Sociology.

OG: Nice!

TL: Social Change.

OG: Interesting!

TL: And Statistics.

OG: Yikes! I’m sorry. That must be horrible for you.

TL: No, I love it.

OG: Really? (at this point, OG usually tilts her head to one side and squints a little). Well, I hated my statistics course when I was in school . . .

Let’s get one thing out in the open right away: statistics has a bad reputation. Though I’m not a fan of speaking in odds (more on that much later in the book), I would bet that odds are good you are not thrilled to be sitting in front of a book on statistics. Other emotions likely are in play: boredom, trepidation, fear. Maybe not all of these, but if you’re like many students taking a course in statistics, the probability is high that some of these emotions are involved. Any effort I make here to dispel such emotions likely will elicit another set of reactions: skepticism, disbelief, anger. I realize it might take me a while to win you over. But I will do my best. I’d even say that the odds are high that at some point, perhaps not right away, but somewhere down the road, you will, perhaps secretly, start to like statistics.

OK, you may not get to that point. But I do hope to convince you that understanding statistics is completely possible if you have the right combination of guides (your instructor and me). It is not only possible to understand statistics; it is also absolutely essential to being an informed and effective student, citizen, activist, or employee. We live in an age in which information is overwhelmingly everywhere, and a lot of this information is statistical. Legislators measure the success of social policies based on statistics. A philanthropist considering a large donation to a nonprofit organization may ask for evidence of the organization’s prior success, and this evidence is often statistical in nature. Start-up companies have made fortunes by developing better statistical models to help people mine the data created daily by people’s Internet searches and by consumer behavior (a journalist even went so far as to call these people “the Numerati” (Baker 2008)). Therefore, if you can’t speak statistics, or read them, you could very well be left out of all of these loops.

Did I just say, “speak statistics”? Yes, I did. In many ways, for many people, learning statistics is very similar to learning a foreign language. If I started speaking, say, Farsi or Swahili right now, I’d probably lose your interest rather quickly (unless, of course, you’re a speaker of these languages, in which case you’d probably perk right up). But do I lose you any less slowly when I say “Adding the squared age term raises the explained variation by 0.04 (with an F-test significant at p < .01) and causes the slope for the interaction effect to lose its statistical significance?” I’d bet not. Right now, to figure out what this sentence meant, you’d need to take it to someone who speaks statistics, and you’d be relying on that person’s translation. By the end of this book, you’ll be able to figure out on your own what such sentences mean, which portends that, among your friends, family, and coworkers, you will likely become the statistical translator. And those statistical tables you see in academic journals or policy briefings? You know, those tables that you just skip over because you have no idea what they’re saying? I’ll give you the necessary skills to be able to navigate such tables with ease.

This book differs substantially from other introductory statistics books. I think that’s a good thing, but, granted, I’m biased. In addition to using a writing style I hope will not bore or confuse you, I get us through the basic statistics relatively quickly. I do this in order to spend much more time than most books do on the statistical techniques that are used most in the real world. In my opinion, many books spend far too many chapters going over statistical techniques that students likely will never see in practice. Then, before they get to the really good stuff, the book ends. This is akin to a movie that has lots of character and plot development, and then, right at the climax, when the school bus filled with orphans is hanging off the cliff, the screen fades to black and the credits roll. This book, in contrast, not only saves those orphans; it finds them all families and buys each child a puppy. In this book, I cover the basics and then get to the good stuff. Although I’ve done my best to write as clearly as possible, there inevitably will be points where, the first time you read through them, something just doesn’t make sense. During such moments, don’t give up right away. Sometimes this material takes a few readings before you really understand it. But, if you are persistent, you will get there.

What Data Look Like

Yes, look. The word data is the plural form of the singular word datum. It may sound weird now, but get used to it, because it’s grammatically correct. Stratum, medium, datum; strata, media, data. The data are correct. The data are available on the Internet. The data do not lie. Actually, sometimes they do lie, but more on that later in the book. In our trip together, we’ll be calculating and interpreting statistics using lots and lots of data, so the first things I want to go over with you are the basic forms that data take, the major sources of data today, and some useful ways to work with data to answer the questions you want to answer.

Most, though not all, quantitative social research data start with surveys. A survey interviewer collects data from a survey respondent. Next, that respondent’s answers are translated into numerical codes that the researchers then input into a dataset. The researchers then use the dataset and a statistical program to calculate their statistics. Reducing people’s complex behaviors and attitudes to numbers is not a perfect process. Interesting details sometimes get lost in translation. I’ll be the first to defend those who use more qualitative techniques to study the social world. However, because this is a book on statistics, we’ll be working with the more quan...

Table of contents

• Cover
• Title
• Copyright
• Contents
• Detailed Contents
• Preface
• Acknowledgments
• Chapter 1: Life in a Data-Laden Age: Finding and Managing Datasets
• Chapter 2: The Art of Visual Storytelling: Creating Accurate Tables and Graphs
• Chapter 3: Summarizing Center and Diversity: Basic Descriptive Statistics
• Chapter 4: Using Sample Crosstabs to Talk about Populations: The Chi-Square Test
• Chapter 5: Using a Sample Mean or Proportion to Talk about a Population: Confidence Intervals
• Chapter 6: Using Multiple Sample Means to Talk about Populations: T-Tests and ANOVA
• Chapter 7: Give Me One Good Reason Why: Bivariate Correlation and Regression
• Chapter 8: Using Sample Slopes to Talk about Populations: Inference and Regression
• Chapter 9: It's All Relative: Dichotomies as Independent Variables in Regression
• Chapter 10: Above and Beyond: The Logic of Controlling and the Power of Nested Regression Models
• Chapter 11: Some Slopes are Bigger than Others: Calculating and Interpreting Beta Coefficients
• Chapter 12: Different Slopes for Different Folks: Interaction Effects
• Chapter 13: Explaining Dichotomous Outcomes: Logistic Regression
• Chapter 14: Visualizing Causal Stories: Path Analysis
• Chapter 15: Questioning the Greatness of Straightness: Nonlinear Relationships
• Chapter 16: Problems and Prospects: Regression Diagnostics, Advanced Techniques, and Where to Go Now
• Appendix A: Variables and Indexes from the Datasets Used in the End-of-Chapter Exercises: Codebooks of the variables used in the SPSS demonstrations and end-of-chapter exercises, explanations of the dozens of indexes created from these variables
• Appendix Β: 100 Articles That Use Statistics in Less than Scary Ways: Descriptions of 100 fascinating but relatively straightforward articles from sociology, political science, criminology, public health, and business; plus, for each article, confusing things to watch out for, and a few questions to consider
• Appendix C: Statistical Tables
• Appendix D: Answers to Odd-Numbered End-of-Chapter Exercises
• Bibliography
• Glossary/Index