Contents & Introduction

Preface to the Fourth Edition

Choosing an Appropriate Statistical Procedure

Where to Find Things in SPSS

1. Introduction

2. Using Windows

3. Loading Data and Printing

4. General Descriptive Statistics

5. Correlation Coefficients

6. Chi-square Tests

7. Independent-samples, Paired-samples, and One-sample t Tests

8. Mann–Whitney U and Wilcoxon Matched-pairs Tests

9. One-way Analysis of Variance

10. Multifactorial Analysis of Variance

11. Repeated-measures Analysis of Variance

12. Multiple Regression

13. Log-linear Analysis

14. Factor Analysis

15. Charts and Graphs

16. Handling Variables and Large Data Files

17. Syntax Windows

Appendix 1: Handling Dates

Appendix 2: Exporting and Importing Excel Files

References

Index

 

Preface to the Fourth Edition

With the help of this Crash Course, you should be able to learn SPSS quickly and painlessly, provided that you have some background knowledge of statistics. SPSS is not hard to use, and we can explain the basics to you without fuss. In our experience, busy people dislike spending large amounts of time learning computer applications. We believe that most SPSS manuals are far more cumbersome than they need to be. Learning SPSS with more conventional manuals is time-consuming and quite an ordeal.

This book is designed to make things quicker and easier. It grew out of a specific need, and it proved popular because it filled a gap in the market, although since the first edition, some flattering imitations have appeared in print. Almost all computational examples in our Crash Course are taken from real data in published research, rather than hypothetical examples such as are found in most statistics and computing books, but we have chosen small data sets to spare you the time and boredom involved in inputting data.

The contents and presentation of the book were greatly improved by usability trials that we carried out for the first edition. We sent a rough draft of the book to 15 students and academics at a dozen different universities, all of whom had expressed a wish to learn SPSS but had no previous knowledge or experience of it, and we asked them to work through the course carefully, making notes of everything that they found unclear or felt could be improved, and keeping a record of the time taken to complete the course. The results were enormously helpful. Our readers came up with comments, criticisms, and useful suggestions for every chapter. These responses enabled us to produce a revised version incorporating a vast number of improvements, big and small, and we know of no other SPSS manual that has had the benefit of such systematic feedback from the end-users for whom it is intended. The time taken to complete the course in the usability trials ranged from five and a half to nine hours, with a mean of just under seven hours (6 hours 52 minutes, to be exact), usually spread over several sessions. The content has expanded slightly since then, but most readers should still be able to complete the course within about 10 hours.

The first two chapters are written with complete beginners in mind. They describe the basic features of SPSS and explain from the very beginning how to get it up and running. If you already have some familiarity with Windows-based applications, then we suggest that you just skim these introductory chapters, but do make a note of the less familiar information in sections 2.3 and 2.4. Chapter 3 describes how data are loaded and printed in SPSS, and this will also be fairly familiar territory to many readers. The remaining chapters describe the most widely used statistical techniques and graphic facilities available in SPSS.

Most of the procedures covered in this book are included in the SPSS Base System. The exceptions are repeated-measures analysis of variance (chapter 11) and log-linear analysis (chapter 13), both of which are supplied with the SPSS Advanced Models add-on module, which has to be purchased separately. If you don’t have the Advanced Models module, then you’ll have to skip those chapters.

The first edition of the Crash Course, published in 2000, was designed for use with versions 8 and 9 of SPSS for Windows. The second edition, published in 2003, was for versions 10 and 11, and the third edition, published in 2006, was for versions 10 to 13. The changes for the second and third editions were largely matters of detail, but there were many of them. In almost every paragraph, small alterations had to be made to accommodate changes in SPSS from earlier versions. Because we are very explicit about exactly which keys to press, even minor alterations necessitated textual changes. For the second edition, in response to requests from readers, we also added two completely new chapters, chapter 13 on log-linear analysis and chapter 14 on factor analysis, and for the third edition, we added chapter 16 on handling variables and large data files, chapter 17 on syntax windows, and a short appendix on exporting and importing Excel files.

This fourth edition became necessary because of further minor modifications introduced in SPSS versions 14, 15, and 16. The procedures themselves have remained largely unchanged, but various alterations to the Data Editor, Output Viewer, Chart Editor, and dialog boxes mean that a user running SPSS 14, 15, or 16 cannot always follow the key strokes precisely as set out in earlier editions. We have deleted some very elementary material on using Windows, because Windows applications are now so well known. We’ve added a brief comment on splitting files to chapter 4 and, at the suggestion of a colleague, a section on partial correlations to chapter 5. We’ve added lots of useful SPSS procedures, including sorting, classifying, and coding data, inserting variables and cases, and paneling charts and graphs. Throughout the book, we’ve rewritten passages to improve clarity and readability.

The earlier editions of this Crash Course were well received by readers, many of whom have been in touch with us, and there’s been a steady demand for it throughout the English-speaking world. But there’s always room for further improvement, and we believe that this edition represents a significant leap forward.

We’re grateful to everyone who took part in the usability trials, and to others who have offered technical advice and help of various kinds. In particular, we wish to express our gratitude to Joseph Amoah-Nyako, John Armstrong, John Beckett, Sarah Bird, Mark Bowers, Kenneth Cowley, Simon Dunkley, Joanne Emery, Sarah Fishburn, Gerry Gardner, Kate Garland, Erica Grossman, Rob Hemmings, Richard Joiner, Geoff Lowe, Sandy MacRae, Rhonda Pearce, Ian Pountney, Caroline Salinger, Berni Simmons, Kathy Smith, Helga Sneddon, Jonathan Stirk, David Stretch, Catherine Sugden, Johnny Sung, Carolyn Tarrant, Cathy Thorp, Gary van Heerden, Stephen L. White, Sue Wilson, and Alison Wray. We also wish to acknowledge the support of the University of Leicester in granting us study leave, during which we prepared the latest edition of this book.

We’ve made the book as straightforward as possible, but not totally idiot-proof, partly because that wouldn’t have been possible and partly because only an idiot would want to read an idiot-proof book. But we’ve done our best to make it clear, explicit, and user-friendly, and we’d appreciate hearing from students and researchers about any further improvements that might be worth introducing into future editions. We’ll acknowledge everyone who offers helpful suggestions unless they ask us not to. Feel free to e-mail us directly about the contents or presentation of the book, or write to us care of the publisher, but please don’t ask us for statistical help or advice on how to analyse your data.

Andrew Colman (amc@le.ac.uk)
Briony Pulford (bdp5@le.ac.uk)


Introduction

When SPSS Inc. of Chicago, Illinois, USA was founded in 1968, the letters SPSS stood for Statistical Package for the Social Sciences. Later, as the company grew beyond its purely academic roots, this was changed to Statistical Product and Service Solutions. Today, the company uses SPSS as a name and no longer as an abbreviation for something else. The detailed operations described in this book apply specifically to versions 14, 15, and 16, and the screenshots are from version 15. For earlier versions of SPSS, there are slight variations, but most of the essential features remain the same. In version 16, most of the buttons are the same as in versions 14 and 15, but some are transposed, so the ones that appear in the screenshots on the right-hand side are now along the bottom, and those on the bottom are now down the right.

SPSS is the oldest and most popular of the many packages of computer programs currently available for statistical analysis. Although it’s extremely powerful, it’s relatively easy to use once you’ve been taught the rudiments. We can teach you the rudiments quite quickly, and you’ll certainly need our guidance, because the package is not self-explanatory and you cannot simply teach yourself to use it just by fiddling around and using the help menu, as one of us was annoyed to discover long ago. For both of us, and many people we’ve spoken to, the chief problem in learning to use it is that the various manuals on the market – some issued by SPSS Inc. and many more by independent writers – are too detailed, too complicated, and above all too long to provide the quick introduction that we need. This book is aimed at readers like ourselves who lack the time to plough through thick manuals, or the patience to submit to a screen-based tutor, but who want to be able to pick up the essential skills for performing standard statistical analyses with SPSS, and who prefer to learn these skills rapidly and painlessly. If you are one of those people who are happy to spend many evenings and weekends learning SPSS the long way, then our considered advice to you is that you should get out more and develop some new leisure activities.

Chapter 2 will focus on the essential information that you need for getting started. If you’re already familiar with Windows, then you only really need to read sections 2.3 and 2.4. The chapters that follow will tell you how to load data, how to print results, how to obtain descriptive statistics, including means, standard deviations, and variances, how to compute Pearson’s correlation coefficient, partial correlations, Spearman’s rho, chi-square tests, t tests for independent and paired samples, Mann–Whitney U tests, Wilcoxon matched-pairs tests, analysis of variance in all its major forms, multiple regression, log-linear analysis, and factor analysis, how to draw charts and graphs, how to change and create variables, how to handle data files, and how to work with SPSS syntax windows. The statistical procedures covered by this book include the most important ones used by psychologists and other social and behavioural scientists. Once you’ve mastered these techniques, you should have little difficulty teaching yourself other procedures available in SPSS.

This book will not teach you statistics. We assume that you already know enough about statistics to understand what assumptions are made about the data that you enter into SPSS, what procedures to use for analysing the data, and how to interpret the results. There’s no point trying to analyse data unless you know what you’re doing. If you need to brush up on your statistics, there are many good books for you to consult. Among the ones that we’re happy to recommend are Hays (2007), Howell (2008), Huck (2008), Norman and Streiner (2008), and Pagano (2007). (Bibliographical details can be found in the list of references at the back of this book.) We have, none the less, included very brief introductions to the essential ideas behind the statistical procedures at the beginning of most chapters, and in the preliminary pages there’s a flow chart to help you choose an appropriate statistical procedure and a table showing where to find things in SPSS. The flow chart and table are restricted to the most commonly used procedures specifically dealt with in this book. There are far more statistical procedures available in SPSS, and both the flow chart and the table are only rudimentary, in the spirit of the book as a whole.

Even if you know what you’re doing, the output that you obtain will be of little value if your data are of poor quality. This nugget of truth is expressed in the computer slang word gigo, which stands for garbage in, garbage out. Awesome though it is, SPSS is not a magic oven that can miraculously transform garbage input into haute cuisine output. To get useful output, you need properly collected data and carefully considered statistical analysis.

We hope and expect that this book will put you on the road to becoming a fluent and efficient SPSS data analyst. Believe it or not, data analysis is fun, once you get the hang of it. Our usability trials, referred to in the preface, suggest that our Crash Course in SPSS for Windows should not take more than about 10 hours and that most people find it quite enjoyable. Happy computing!