Working With Numbers and Statistics

A Handbook for Journalists

Charles Livingston, Paul S. Voakes

• Publisher: Routledge
• Year: 2005
• eBook ISBN: 9781135605933
• Topic: Languages & Linguistics
• Subtopic: Communication Studies
• Index: Languages & Linguistics

Working With Numbers and Statistics: A Handbook for Journalists will bolster math skills and improve math confidence for journalists at all skill levels. Authors Charles Livingston and Paul Voakes developed this resource book to improve journalistic writing and reporting, enabling journalists to:
*make accurate, reliable computations, which in turn enables one to make relevant comparisons, put facts into perspective, and lend important context to stories;
*recognize inaccurate presentations, whether willfully spun or just carelessly relayed;
*ask appropriate questions about numerical matters;
*translate complicated numbers for viewers and readers in ways they can readily understand;
*understand computer-assisted reporting; and
*write livelier, more precise pieces through the use of numbers.

The math is presented in a journalistic context throughout, enabling readers to see how the procedures will come into play in their work.

Working With Numbers and Statistics is designed as a reference work for journalism students developing their writing and reporting skills. It will also serve professionals as a useful tool to improve their understanding and use of numbers in news stories.

chapter
1
Introduction

Numbers, numbers, numbers. There’s just no avoiding them, especially when you’re a journalist. Numbers aren’t just on the sports or financial pages. Whether you’re reporting on local tax rates, medical research reports, school district budgets, environmental impact reports, box-office receipts, or any other subject most journalists consider newsworthy, the odds are good that you’ll encounter numbers. And the odds are also good that you can tell the story better with the appropriate (and accurate) use of numbers.

This book had its origins in a course called Mathematics and Statistics for Journalism, at Indiana University. By 1999 the journalism faculty had begun to realize that their students, even when required to take math and statistics elsewhere on campus, were not applying any math skills to their reporting, writing, or editing classes. Meanwhile, Indiana’s Department of Mathematics was coordinating a campuswide effort, funded by the National Science Foundation, to create math courses specific to other disciplines. Journalism was a natural for this project, and your two co-authors set about designing and teaching a course specifically for journalism majors.

We discovered in our students the same reluctance (some call it phobia) regarding math that we’ve seen among professional journalists over the years. After all, isn’t this why many journalists became writers and editors in the first place, be-cause science and math caused them to break out in hives? Journalism students may not have realized it yet, but most professional journalists do recognize, perhaps grudgingly, how math skills make you a better journalist:

• You can make accurate, reliable computations, which in turn enable you to make relevant comparisons, put facts into perspective and lend important context to stories.
• You can protect yourself from being fooled by inaccurate presentations of data, whether willfully spun or just carelessly relayed. If we value independence as a cornerstone of journalism ethics, part of that independence must be the ability to assess numerical information without relying on the source.
• You can ask appropriate questions about numerical matters.
• You can translate complicated numbers for your viewers and readers, in ways they can readily understand. But it takes a full understanding on your part first.
• You can hold your own in the brave new world of computer-assisted reporting, which is no longer the preserve of newsroom geeks.
• Your writing can actually become livelier, because numbers give you precision, and precision always improves writing. Consider the following statements:

• “My husband is a moderate smoker.”
• “I’ve read several books this summer.”
• The mayor has a comfortable income.”

So how many cigarettes a day does the husband smoke? How many books read? How much does the mayor make?

As simple and obvious as it may seem, numbers convey important meaning. But too often journalists skirt around the numerical parts of their fact-gathering and informationsharing, because they think they’re incompetent with numbers.

This little book is designed to bolster a journalist’s math skills, and in turn bolster a journalist’s math confidence. We’ll start with some math basics. It’s been years, maybe decades, since you last computed a percentage change or found an average. For example, in seventh grade you were probably adept at figuring out the percentage change when a population has risen from 41,000 to 49,000. Every journalist should feel confident doing fairly simple procedures like that (and that’s right; it was a 19.5 percent increase).

We’ll also cover the basics of statistics, called descriptive statistics: the mathematical procedures people use to summarize vast amounts of data. For example, if Republicans claim the average salary of state workers is $37,000 and Democrats claim the average worker earns only $31,000, is one party lying? Not necessarily: One “average” could be the mean, and the other figure could be the median. Every journalist should know the difference between the two, and when sources might be tempted to use one or the other to their advantage.

Then we’ll move slightly beyond the basics, to inferential statistics—the procedures people use to draw conclusions from the initial statistical findings. Even if they never compute statistics on their own, journalists need to know the basic language and logic of statistics. For example, if the latest poll has one candidate with 49 percent support of those likely to vote, and the other candidate with 47 percent, can we conclude that the first candidate holds a slim lead? No—at least not without a lot more information. Another example: If a national health survey finds that middle-aged men on the West Coast and Hawaii had less heart disease than middle-aged men in the Midwest and South, does that mean we should advise our viewers to go west (if they want to live longer)? No—there may be a more pertinent reason than geography for that difference.

We’ll review some of the little things as well, like usage and style when you’re writing with numbers. For example, which of the following is better writing?

The city council approved a budget increase of $39.2 million.

The city council approved a budget increase of 17 percent.

It’s the second sentence, because the quantity 17 percent has more immediate meaning to readers than the quantity $39.2 million, especially with no more context than provided.

We’ve tried to provide the same basic structure in each of the chapters that follow. We’ll always start with the math, so that if all you need is a reminder of something you learned long ago, you can plunge straight into it. If the key mathematical procedure is a fundamental formula or a rule of thumb, you’ll find that displayed in a box. For every procedure we’ll provide an example or two, to show how the math would be done in a specific situation. At the end of every section, we’ll put the math into a more journalistic context, so you can see how the procedures are likely to come into play in the work of a reporter or editor.

We’ve tried to keep the statistics simple, telling you what you need to know to deal with the situations journalists are most likely to confront. But we’ve also added a chapter called Advanced Statistics, which elaborates on the basics from chapters 4 and 5.

This book is not designed as a math or stats textbook—there are plenty of those floating around college math departments—so you won’t see practice problem sets or elongated examples. Rather we hope you’ll keep this as a reference work for your bookshelf, right alongside your style book, dictionary, almanac and whatever other works you keep handy for daily use.

Without skills in math or stats, journalists constantly have to rely on the calculations and interpretations of their sources, and constantly hope and pray that the numbers they use in their writing are appropriate and correct. That situation presents a picture of neither independence nor accuracy. However, journalists armed with some logic, technique, and interpretive skills can analyze research, ask appropriate questions and understand the data well enough to tell readers and viewers clearly what the numbers mean. We hope this book can help you achieve those goals.

---

Summary of Chapters

Chapter 2 presents basic mathematical concepts; we haven’t tried to give a review course in elementary mathematics, but rather have identified some of the topics that come up most frequently for journalists. Chapter 2 also presents some of the basic mathematics that arises in working with budgets, taxes, and in business reporting. Chapters 3, 4, 5, and 6as are the heart of our presentation on statistics: Chapter 3 tells you some of the tools used to describe data; Chapter 4 tells you how to use data to estimate unknown numbers; Chapter 5 describes how statistical methods permit you to draw inferences from those estimates and judge the reliability of those inferences; Chapter 6 describes surveying techniques, as surveys are the most important method used to gather data about large groups. Chapter 7 presents advanced topics in statistics, those that underlie the previous material and might arise occasionally for the journalist. In Chapter 8 each section is focused on a common misunderstanding that arises in the use of statistical methods, including an extended section on the contrast between “probability” and “odds.” The concluding chapters are of independent interest: Chapter 9 summarizes the usefulness of computers to mathematical and statistical work of journalists: resources on the Internet and the use of Microsoft Excel in working with data; Chapter 10 focuses on the writing aspects of working with numbers.

In the text we use special icons to indicate examples and warnings:

This icon denotes basic examples, often with mathematical content.

This icon marks examples that are focused on writing issues.

This icon indicates a warning.

chapter
2
Some Fundamentals

2.1. Percentages

---

Computing Percentages

To convert a percentage to a (decimal) number, shift the decimal point two places to the left.

15 percent is the same as .15.

To convert a decimal to the percentage, shift the decimal point two places to the right and add “percent.”

.49 is 49 percent.

To compute a certain percentage of a certain number, multiply the number by the percentage in decimal form.

To compute 75 percent of 200, multiply 200 by 75 percent, or 75, and you get 150. So 150 is 75 percent of 200.

To compute a percentage when all you have is two numbers, divide the “portion” by the “whole.”

What percentage of 200 is represented by 30? Here 30 is the portion, and 200 is the whole. Dividing 30 by 200 yields .15, or 15 percent.

Sometimes the “whole” can actually be smaller than the portion,” but the same rule applies.

What percentage of 200 is represented by 300? Here 200 is again the “whole” (because we want the percentage “of 200”), so divide 300 by 200 and you get 1.5. Shift the decimal point two places to the right and you get 150 percent.

---

Percentage Change

By how much did a budget, or a batting average, or a company’s profit, go up or down? To calculate percentage changes, you need to focus on two key amounts: the amount of change, and the “old number” amount.

To compute a percentage change, divide the amount of change by the “old number.”

The price of ZippyCar in 2002 was $18,000. In 2003 the price had increased to $19,200. What was the percentage increase? First, determine the amount of change and the “old number.” In this example the “old number” is $18,000. To get the amount of change, subtract $18,000 from $19,200 and you get $1,200. Divide the amount of change by the “old number” ($1,200 divided by $18,000) and you get .067, or 6.7 percent. The price of the car had risen by nearly percent.

The price of a SnazzyCar...

Table of contents

• Cover Page
• Lea’s Communication Series
• Title Page
• Copyright Page
• Preface
• Chapter 1: Introduction
• Chapter 2: Some Fundamentals
• Chapter 3: Describing Data
• Chapter 4: Estimation
• Chapter 5: Inference: Drawing Conclusions from Data
• Chapter 6: Surveying and Experimental Design
• Chapter 7: Advanced Statistics
• Chapter 8: Cautions and Fallacies
• Chapter 9: Excel and the Internet
• Chapter 10: Writing with Numbers
• Appendix