Practical Statistics for Geographers and Earth Scientists

Nigel Walford has taught Geographical Information Systems courses at Kingston University for 12 years and previously worked in the Data Archive at Essex University. He graduated from the University of Sussex and did his postgraduate work at the University of London. His thematic research interests relate mainly to rural issues and include agriculture, land use planning and population. Recent book publications include Geographical Data Analysis and Reshaping the Countryside: Perceptions and Processes of Rural Change as well as several journal articles in Health and Place, Geography, Planning Practice and Research, Applied Geography and International Journal of Population Geography.

Preface xi

Acknowledgements xiii

Glossary xv

Section 1 First principles 1

1 What's in a number? 3

Learning outcomes

1.1 Introduction to quantitative analysis 4

1.2 Nature of numerical data 9

1.3 Simplifying mathematical notation 14

1.4 Introduction to case studies and structure of the book 19

2 Geographical data: quantity and content 21

Learning outcomes

2.1 Geographical data 21

2.2 Populations and samples 22

2.3 Specifying attributes and variables 43

3 Geographical data: collection and acquisition 57

Learning outcomes

3.1 Originating data 58

3.2 Collection methods 59

3.3 Locating phenomena in geographical space 87

4 Statistical measures (or quantities) 93

Learning outcomes

4.1 Descriptive statistics 93

4.2 Spatial descriptive statistics 96

4.3 Central tendency 100

4.4 Dispersion 118

4.5 Measures of skewness and kurtosis for nonspatial data 124

4.6 Closing comments 129

5 Frequency distributions, probability and hypotheses 131

Learning outcomes

5.1 Frequency distributions 132

5.2 Bivariate and multivariate frequency distributions 137

5.3 Estimation of statistics from frequency distributions 145

5.4 Probability 149

5.5 Inference and hypotheses 165

5.6 Connecting summary measures, frequency distributions and probability 169

Section 2 Testing times 173

6 Parametric tests 175

Learning outcomes

6.1 Introduction to parametric tests 176

6.2 One variable and one sample 177

6.3 Two samples and one variable 201

6.4 Three or more samples and one variable 210

6.5 Confi dence intervals 216

6.6 Closing comments 219

7 Nonparametric tests 221

Learning outcomes

7.1 Introduction to nonparametric tests 222

7.2 One variable and one sample 223

7.3 Two samples and one (or more) variable(s) 245

7.4 Multiple samples and/or multiple variables 256

7.5 Closing comments 264

Section 3 Forming relationships 265

8 Correlation 267

Learning outcomes

8.1 Nature of relationships between variables 268

8.2 Correlation techniques 275

8.3 Concluding remarks 298

9 Regression 299

Learning outcomes

9.1 Specification of linear relationships 300

9.2 Bivariate regression 302

9.3 Concluding remarks 336

10 Correlation and regression of spatial data 341

Learning outcomes

10.1 Issues with correlation and regression of spatial data 342

10.2 Spatial and temporal autocorrelation 345

10.3 Trend surface analysis 378

10.4 Concluding remarks 394

References 397

Further Reading 399

Index 403

Plate section: Statistical Analysis Planner and Checklist falls between pages 172 and 173