bioestadistica, Apuntes de Bioestadística

¡Descarga bioestadistica y más Apuntes en PDF de Bioestadística solo en Docsity!

CHAPMAN & HALL/CRC

A CRC Press Company Boca Raton London New York Washington, D.C.

Sabine Landau

and

Brian S. Everitt

A Handbook of

Statistical Analyses

using SPSS

Preface

SPSS, standing for Statistical Package for the Social Sciences , is a powerful, user-friendly software package for the manipulation and statistical analysis of data. The package is particularly useful for students and researchers in psychology, sociology, psychiatry, and other behavioral sciences, contain- ing as it does an extensive range of both univariate and multivariate procedures much used in these disciplines. Our aim in this handbook is to give brief and straightforward descriptions of how to conduct a range of statistical analyses using the latest version of SPSS, SPSS 11. Each chapter deals with a different type of analytical procedure applied to one or more data sets primarily (although not exclusively) from the social and behav- ioral areas. Although we concentrate largely on how to use SPSS to get results and on how to correctly interpret these results, the basic theoretical background of many of the techniques used is also described in separate boxes. When more advanced procedures are used, readers are referred to other sources for details. Many of the boxes contain a few mathematical formulae, but by separating this material from the body of the text, we hope that even readers who have limited mathematical background will still be able to undertake appropriate analyses of their data. The text is not intended in any way to be an introduction to statistics and, indeed, we assume that most readers will have attended at least one statistics course and will be relatively familiar with concepts such as linear regression , correlation , significance tests, and simple analysis of variance. Our hope is that researchers and students with such a background will find this book a relatively self-contained means of using SPSS to analyze their data correctly. Each chapter ends with a number of exercises, some relating to the data sets introduced in the chapter and others introducing further data sets. Working through these exercises will develop both SPSS and statistical skills. Answers to most of the exercises in the text are provided at

http://www.iop.kcl.ac.uk/iop/departments/BioComp/SPSSBook.shtml. The majority of data sets used in the book can be found at the same site. We are grateful to Ms. Harriet Meteyard for her usual excellent word processing and overall support during the writing of this book.

Sabine Landau and Brian Everitt London, July 2003

Contents

Preface Distributors

1 A Brief Introduction to SPSS

1.1 Introduction 1.2 Getting Help 1.3 Data Entry 1.3.1 The Data View Spreadsheet 1.3.2 The Variable View Spreadsheet 1.4 Storing and Retrieving Data Files 1.5 The Statistics Menus 1.5.1 Data File Handling 1.5.2 Generating New Variables 1.5.3 Running Statistical Procedures 1.5.4 Constructing Graphical Displays 1.6 The Output Viewer 1.7 The Chart Editor 1.8 Programming in SPSS

2 Data Description and Simple Inference for Continuous

Data: The Lifespans of Rats and Ages at Marriage in the U.S. 2.1 Description of Data 2.2 Methods of Analysis. 2.3 Analysis Using SPSS 2.3.1 Lifespans of Rats 2.3.2 Husbands and Wives 2.4 Exercises 2.4.1 Guessing the Width of a Lecture Hall 2.4.2 More on Lifespans of Rats: Significance Tests for Model Assumptions 2.4.3 Motor Vehicle Theft in the U.S. 2.4.4 Anorexia Nervosa Therapy 2.4.5 More on Husbands and Wives: Exact Nonparametric Tests

3 Simple Inference for Categorical Data: From Belief in

the Afterlife to the Death Penalty and Race 3.1 Description of Data 3.2 Methods of Analysis 3.3 Analysis Using SPSS 3.3.1 Husbands and Wives Revisited. 3.3.2 Lifespans of Rats Revisited 3.3.3 Belief in the Afterlife 3.3.4 Incidence of Suicidal Feelings 3.3.5 Oral Contraceptive Use and Blood Clots 3.3.6 Alcohol and Infant Malformation 3.3.7 Death Penalty Verdicts 3.4 Exercises 3.4.1 Depersonalization and Recovery from Depression 3.4.2 Drug Treatment of Psychiatric Patients: Exact Tests for Two-Way Classifications 3.4.3 Tics and Gender 3.4.4 Hair Color and Eye Color

4 Multiple Linear Regression: Temperatures in America

and Cleaning Cars 4.1 Description of Data 4.2 Multiple Linear Regression 4.3 Analysis Using SPSS 4.3.1 Cleaning Cars 4.3.2 Temperatures in America 4.4 Exercises. 4.4.1 Air Pollution in the U.S. 4.4.2 Body Fat 4.4.3 More on Cleaning Cars: Influence Diagnostics

5 Analysis of Variance I: One-Way Designs; Fecundity of

Fruit Flies, Finger Tapping, and Female Social Skills. 5.1 Description of Data 5.2 Analysis of Variance. 5.3 Analysis Using SPSS 5.3.1 Fecundity of Fruit Flies. 5.3.2 Finger Tapping and Caffeine Consumption. 5.3.3 Social Skills of Females 5.4 Exercises. 5.4.1 Cortisol Levels in Psychotics: Kruskal-Wallis Test 5.4.2 Cycling and Knee-Joint Angles 5.4.3 More on Female Social Skills: Informal Assessment of MANOVA Assumptions

10 Survival Analysis: Sexual Milestones in Women and

Field Dependency of Children. 10.1 Description of Data 10.2 Survival Analysis and Cox’s Regression 10.3 Analysis Using SPSS 10.3.1 Sexual Milestone Times 10.3.2 WISC Task Completion Times 10.4 Exercises 10.4.1 Gastric Cancer 10.4.2 Heroin Addicts 10.4.3 More on Sexual Milestones of Females

11 Principal Component Analysis and Factor Analysis:

Crime in the U.S. and AIDS Patients’ Evaluations of Their Clinicians 11.1 Description of Data 11.2 Principal Component and Factor Analysis 11.2.1 Principal Component Analysis 11.2.2 Factor Analysis 11.2.3 Factor Analysis and Principal Components Compared 11.3 Analysis Using SPSS 11.3.1 Crime in the U.S. 11.3.2 AIDS Patients’ Evaluations of Their Clinicians 11.4 Exercises 11.4.1 Air Pollution in the U.S. 11.4.2 More on AIDS Patients’ Evaluations of Their Clinicians: Maximum Likelihood Factor Analysis

12 Classification: Cluster Analysis and Discriminant

Function Analysis; Tibetan Skulls 12.1 Description of Data 12.2 Classification: Discrimination and Clustering 12.3 Analysis Using SPSS 12.3.1 Tibetan Skulls: Deriving a Classification Rule. 12.3.2 Tibetan Skulls: Uncovering Groups. 12.4 Exercises 12.4.1 Sudden Infant Death Syndrome (SIDS) 12.4.2 Nutrients in Food Data 12.4.3 More on Tibetan Skulls

References

Chapter 1

A Brief Introduction

to SPSS

1.1 Introduction

The “Statistical Package for the Social Sciences” (SPSS) is a package of programs for manipulating, analyzing, and presenting data; the package is widely used in the social and behavioral sciences. There are several forms of SPSS. The core program is called SPSS Base and there are a number of add-on modules that extend the range of data entry, statistical, or reporting capabilities. In our experience, the most important of these for statistical analysis are the SPSS Advanced Models and SPSS Regression Models add-on modules. SPSS Inc. also distributes stand-alone programs that work with SPSS. There are versions of SPSS for Windows (98, 2000, ME, NT, XP), major UNIX platforms (Solaris, Linux, AIX), and Macintosh. In this book, we describe the most popular, SPSS for Windows, although most features are shared by the other versions. The analyses reported in this book are based on SPSS version 11.0.1 running under Windows 2000. By the time this book is published, there will almost certainly be later versions of SPSS available, but we are confident that the SPSS instructions given in each of the chapters will remain appropriate for the analyses described. While writing this book we have used the SPSS Base, Advanced Models, Regression Models, and the SPSS Exact Tests add-on modules. Other avail- able add-on modules ( SPSS Tables , SPSS Categories , SPSS Trends , SPSS Missing Value Analysis ) were not used.

1.2 Getting Help

Online help is provided from the Help menu or via context menus or Help buttons on dialogue boxes. We will mention the latter features when discussing the dialogue boxes and output tables. Here, we concentrate on the general help facility. The required menu is available from any window and provides three major help facilities:

Help — Statistics Coach helps users unfamiliar with SPSS or the statistical procedures available in SPSS to get started. This facility prompts the user with simple questions in nontechnical language about the purpose of the statistical analysis and provides visual examples of basic statistical and charting features in SPSS. The facility covers only a selected subset of procedures. Help — Tutorial provides access to an introductory SPSS tutorial, includ- ing a comprehensive overview of SPSS basics. It is designed to provide a step-by-step guide for carrying out a statistical analysis in SPSS. All files shown in the examples are installed with the tutorial so the user can repeat the analysis steps. Help — Topics opens the Help Topics: SPSS for Windows box, which pro- vides access to Contents, Index, and Find tabs. Under the Contents tab, double-clicking items with a book symbol expands or col- lapses their contents (the Open and Close buttons do the same). The Index tab provides an alphabetical list of topics. Once a topic is selected (by double-clicking), or the first few letters of the word are typed in, the Display button provides a description. The Find tab allows for searching the help files for specific words and phrases.

1.3 Data Entry

When SPSS 11.0 for Windows is first opened, a default dialogue box appears that gives the user a number of options. The Tutorial can be accessed at this stage. Most likely users will want to enter data or open an existing data file; we demonstrate the former (Display 1.1). Further options will be discussed later in this chapter. This dialogue box can be prevented from opening in the future by checking this option at the bottom of the box. When Type in data is selected, the SPSS Data Editor appears as an empty spreadsheet. At the top of the screen is a menu bar and at the bottom a status bar. The status bar informs the user about facilities currently active; at the beginning of a session it simply reads, “SPSS Processor is ready.”

The facilities provided by the menus will be explained later in this chapter. SPSS also provides a toolbar for quick and easy access to common tasks. A brief description of each tool can be obtained by placing the cursor over the tool symbol and the display of the toolbar can be controlled using the command Toolbars… from the View drop-down menu (for more details, see the Base User’s Guide , SPSS Inc., 2001d).

1.3.1 The Data View Spreadsheet

The Data Editor consists of two windows. By default the Data View, which allows the data to be entered and viewed, is shown (Display 1.2). The other window is the Variable View, which allows the types of variables to be specified and viewed. The user can toggle between the windows by clicking on the appropriate tabs on the bottom left of the screen. Data values can be entered in the Data View spreadsheet. For most analysis SPSS assumes that rows represent cases and columns variables. For example, in Display 1.2 some of five available variable values have been entered for twenty subjects. By default SPSS aligns numerical data entries to the right-hand side of the cells and text (string) entries to the left-hand side. Here variables sex, age, extrover, and car take numerical

Display 1.1 Initial SPSS for Windows dialogue box.

There are 10 characteristics to be specified under the columns of the Variable View (Display 1.3):

1. Name — the chosen variable name. This can be up to eight alphanumeric characters but must begin with a letter. While the underscore (_) is allowed, hyphens (-), ampersands (&), and spaces cannot be used. Variable names are not case sensitive.
2. Type — the type of data. SPSS provides a default variable type once variable values have been entered in a column of the Data View. The type can be changed by highlighting the respective entry in the second column of the Variable View and clicking the three-periods symbol (…) appearing on the right-hand side of the cell. This results in the Variable Type box being opened, which offers a number of types of data including various formats for numerical data, dates, or currencies. (Note that a common mistake made by first-time users is to enter categorical variables as type “string” by typing text into the Data View. To enable later analyses, categories should be given artificial number codes and defined to be of type “numeric.”)
3. Width — the width of the actual data entries. The default width of numerical variable entries is eight. The width can be increased or decreased by highlighting the respective cell in the third column and employing the upward or downward arrows appearing on the

Display 1.3 Variable View window of the Data Editor.

right-hand side of the cell or by simply typing a new number in the cell.

4. Decimals — the number of digits to the right of the decimal place to be displayed for data entries. This is not relevant for string data and for such variables the entry under the fourth column is given as a greyed-out zero. The value can be altered in the same way as the value of Width.
5. Label — a label attached to the variable name. In contrast to the variable name, this is not confined to eight characters and spaces can be used. It is generally a good idea to assign variable labels. They are helpful for reminding users of the meaning of variables (placing the cursor over the variable name in the Data View will make the variable label appear) and can be displayed in the output from statistical analyses.
6. Values — labels attached to category codes. For categorical variables, an integer code should be assigned to each category and the variable defined to be of type “numeric.” When this has been done, clicking on the respective cell under the sixth column of the Variable View makes the three-periods symbol appear, and clicking this opens the Value Labels dialogue box, which in turn allows assign- ment of labels to category codes. For example, our data set included a categorical variable sex indicating the gender of the subject. Clicking the three-periods symbol opens the dialogue box shown in Display 1.4 where numerical code “0” was declared to represent females and code “1” males.
7. Missing — missing value codes. SPSS recognizes the period symbol as indicating a missing value. If other codes have been used (e.g., 99, 999) these have to be declared to represent missing values by highlighting the respective cell in the seventh column, clicking the

Display 1.4 Declaring category code labels.

A data file shown in the Data Editor can be saved by using the commands Save or Save As…. In the usual Windows fashion Save (or from the toolbar) will save the data file under its current name, overwriting an existing file or prompting for a name otherwise. By contrast, Save As always opens the Save Data As dialogue where the directory, file name, and type have to be specified. SPSS supports a number of data formats. SPSS data files are given the extension .sav. Other formats, such as ASCII text (.dat), Excel (.xls), or dBASE (.dbf), are also available. To open existing SPSS data files we use the commands File – Open – Data… from the menu bar (or from the toolbar). This opens the Open File dialogue box from which the appropriate file can be chosen in the usual way (Display 1.6). Recently used files are also accessible by placing the cursor over Recently Used Data on the File drop-down menu and double- clicking on the required file. In addition, files can be opened when first starting SPSS by checking Open an existing data source on the initial dialogue box (see Display 1.1). SPSS can import data files in other than SPSS format. A list of data formats is provided by selecting the down arrow next to the Files of type field (Display 1.6). There are a number of formats including spreadsheet (e.g., Excel, *.xls), database (e.g., dBase, *.dbf), and ACSII text (e.g., *.txt,

Display 1.5 File drop-down menu.

*.dat). Selecting a particular file extension will cause a dialogue box to appear that asks for information relevant to the format. Here we briefly discuss importing Excel files and ASCII text files. Selecting to import an Excel spreadsheet in the Open File box will bring up the Opening File Options box. If the spreadsheet contains a row with variable names, Read Variable Names has to be checked in this box in order that the first row of the spreadsheet is read into variable names. In addition, if there are initial empty rows or columns in the spreadsheet, SPSS needs to be informed about it by defining the cells to be read in the Range field of the Opening File Options box (using the standard spreadsheet format, e.g., B4:H10 for the cells in the rectangle with corners B4 and H10 inclusive). Selecting to open an ASCII text file in the Open File dialogue box (or selecting Read Text Data from the File drop-down directly, see Display 1.5) causes the Text Import Wizard to start. The initial dialogue box is shown in Display 1.7. The Wizard proceeds in six steps asking questions related to the import format (e.g., how the variables are arranged, whether variable names are included in the text file), while at the same time making suggestions and displaying how the text file would be transformed into an SPSS spreadsheet. The Text Import Wizard is a convenient and self- explanatory ASCII text import tool. (Choosing New from the File drop-down menu will clear the Data Editor spreadsheet for entry of new data.)

Display 1.6 Opening an existing SPSS data file.