UT CSTE ACLU SRI an ce WITH A GUIDE TO SPSS | i I | | second édition © | ‘SAGEContents Preface Part 1 An introduction to statistical analysis 1 Variables and their measurement 3 The conceptualization and operationalization of variables 4 Scales of measurement 7 Levels of measurement 8 Univariate, bivariate, and multivariate analysis 11 Descriptive statistics 14 Exercises 15 2 Setting up an SPSS data file 17 Obtaining a copy of SPSS_17 Alltematives to SPSS. 17 Options for data entry in SPSS 18 The SPSS Data Editor 19 Assigning a variable name 21 Setting the data type 22 Setting the data width and decimal places 23 Defining variable labels 23 : Defining value labels 24 Setting missing values 25 Setting the column format and aligament 26 Specifying the level of measurement 27 Generating variable definitions in SPSS 28 ‘The SPSS Viewer window 32 Saving a data file 32 Data entry 33 Cheeking for incorrect values: Data cleaning 35 Summary 35 Exercises 35 Part 2 Descriptive statistics: Graphs and tables 3. The graphical description of data 39 Some general principles 39 Pie graphs 40 Bar graphs 42 Histograms and polygons 44 Interpreting a univariate distribution 46 Graphing two variables 47 Common problems and misuses of graphs 50 Exercises 53vi Contenis: ‘The tabular description of data 55 Listed data tables 55 Simple frequency tables 55 Relative frequency tables: percentages and proportions 57 Cumulative frequency tables 60 Class intervals 61 Percentiles 64 Frequency tables using SPSS. 65 Valid cases and missing values 67 Improving the look of tables 67 Exercises 68 ‘Using tables to investigate the relationship between variables: Crosstabulations 70 Crosstabulations as descriptive statistics 70 ‘Types of data suitable for crosstabulations 72 Crosstabulations with relative frequencies 73 Crosstabulations using SPSS 74 Interpreting a crosstabulation: The pattern and strength of a relationship 75 Interpreting a crosstabulation when both variables are at least ordinal 76 Summary 78 Exercises 78 ‘Measures of association for crosstabulations: Nominal data 81 “Measures of association as descriptive statistics 81 ‘Measures of association for nominal scales 83 Properties of lambda 86 Lambda using SPSS_ 87 Limitations on the use of lambda 90 Standardizing table frequencies 92 Exercises 93 ‘Measures of association for crosstabulations: Ranked data 95 Data considerations 95 Concordant pairs 96 Discordant pairs 97 ‘Measures of association for ranked data 98 Gamma 99 Somers’ 101 Kendall’s tau-b 102 Kendall's tau-c 102 ‘Measures of association using SPSS 102 Summary 107 Exercises 107 Muttivariate analysis of crosstabs: Elaboration 110 Direct relationship 110 Elaboration of crosstabs using SPSS_ 112 Partial gamma 113 Spurious or intervening relationship? 114 Conditional relationship 115 Summary 117 Exercises 118Contents vii Part 3 Descriptive statistics: Numerical measures 9 Measures of central tendency 123 Measures of central tendency 123 The mode 124 The median 125 The mean 126 Choosing a measure of central tendency 128 Measures of central tendency using SPSS: Univariate analysis 129 Measures of central tendency using SPSS: Bivariate and multivariate analysis 132 Summary 133 Exercises 134 10 Measures of dispersion 136 ‘The range 136 The interquartile range 137 The standard deviation 138 Coefficient of relative variation 140 Index of qualitative variation 141 Measures of dispersion using SPSS. 145 Summary 145 Exercises 146 11 The normal curve 147 The normal distribution 147 : Using normal curves to describe a distribution 150 zscores 151 : Normal curves on SPSS_157 Exercises 159 12 Correlation and regression 161 Scatter plots 161 Linear regression 162 Pearson’s product moment correlation coefficient 169 Explaining variance: The coefficient of determination 170 Plots, correlation, and regression using SPSS 172 The assumptions behind regression analysis. 177 Speatman’s rank-order correlation coefficient 179 Spearman’s rho using SPSS 180 Correlation where the independent variable is categorical: Eta 182 Summary 183 Exercises 183 13 Multiple regression 187 Introduction to multiple regression 188 Multiple regression with SPSS. 190 Testing for the significance of the multivariate model 193 Alternative methods for selecting variables in the regression model 193 Stepwise regression 194 Extending the basic regression analysis: Adding categorical independent variables 197 Further extensions to the basic regression analysis: Hierarchical regression 198 The assumptions behind multiple regression 198 Exercises 199viii Contents Part 4 Inferential statistics: Tests for a mean 14 Sampling distributions 203 Random samples 204 The sampling distribution of a sample statistic 205 The central limit theorem 210 Generating random samples using SPSS. 210 Summary 212 Exercises 212 15 Introduction to hypothesis testing and the one sample z-test for a mean 214 Step 1: State the null and alternative hypotheses 217 Step 2: Choose the test of significance 219 Step 3: Describe the sample and derive the p-score 220 Step 4: Decide at what alpha level, if any, the tesult is statistically significant 222 Step 5: Report results 224 What does it mean when we ‘fail to reject the null hypothesis’? 226 ‘What does it mean to ‘reject the null hypothesis”? 226 A two-tail z-test for a single mean 227 The debate over one-tail and two-tail tests of significance 228 A one-tail z-test for a single mean 229 Summary 230 Appendix: Hypothesis testing using critical values of the test statistic 230 Exercises 231 16 ‘The one sample ¢test for a mean 233 The Student's eistribution 233 The one sample /test fora mean 234 The one sample test using SPSS 238 Summary 239 Exercises 240 17 Inference using estimation and confidence intervals 242 ‘The sampling distribution of sample means 242 Estimation 243 ‘Changing the confidence level 247 Changing the sample size 250 Estimation using SPSS 250 Confidence intervals and hypothesis testing 252 Exercises 253 18 The two samples ¢-test for the equality of means 255 Dependent and independent variables 256 The sampling distribution of the difference between two means 257 ‘The two samples t-test for the equality of means 259 The two samples t-test using SPSS 261 Exercises 264 19 The F-test for the equality of more than two means: Analysis of variance 266 The one-way analysis of variance F-test 269 ANOVA using SPSS_ 272 Summary 277 Exercises 279Contents ix 20 The two dependent samples test for the mean difference 280 Dependent and independent samples 280 ‘The two dependent samples test for the mean difference 281 The two dependent samples ¢-test using SPSS. 283 Exercises 286 Part 5 Inferential statistics: Tests for frequency distributions 21 One sample tests for a binomial distribution 291 Data considerations 291 The sampling distribution of sample percentages 292 The z+test for a binomial percentage 293 ‘The z-test for a binomial percentage using SPSS. 295 Estimating a population percentage 297 The runs test for randomness 299 The runs test using SPSS 302 Exercises 303 22, One sample tests for a multinomial distribution 305 ‘The chi-square goodness-of-fit test 305 Chi-square goodness-of-fit test using SPSS. 308 The chi-square goodness-of-fit test for normality 312 Summary 313 Exercises 314 23 The chi-square test for independence 316 The chi-square test and other tests of significance 316 Statistical independence 317 ‘The chi-square test for independence 317 ‘The distribution of chi-square 322 ‘The chi-square test using SPSS_323 Problems with small samples 328 Problems with large samples 329 Appendix: hypothesis testing for two percentages 331 Exercises 333 24 Frequency tests for two dependent samples 335 The McNemar chi-square test for change 335 The McNemar test using SPSS_337 The sign test 338 Summary 340 Exercises 340 Part 6 Inferential statistics: Other tests of significance 25 Rank-order tests for two or more samples 343 Data considerations 343 ‘The rank sum and mean rank as descriptive statistics 344 The z-test for the rank sum for two independent samples 348 Wilcoxon’s rank sum z-test using SPSS 352 The Wilcoxon signed-ranks z-test for two dependent samples 353 The Wilcoxon signed-ranks test using SPSS_ 356x Contents Other non-parametric tests for two or more samples 357 Appendix: the Mann-Whitney U test 358 Exercises 359 26 The t-test for a correlation coefficient 362 ‘The Mest for Pearson’s correlation coefficient 362 Testing the significance of Pearson’s correlation coefficient using SPSS 364 ‘The /test for Spearman’s rank-order correlation coefficient 365 Testing the significance of Spearman’s correlation coefficient using SPSS. 366 ‘Testing for significance in multiple regression 367 Exercises 368 Appendix 369 ‘Table Al Area under the standard normal curve 369 Table A2 Critical values for distributions 370 Table A3 Critical values for F-distributions (a= 0,05) 371 Table A4 Critical values for chi-square distributions 372 Table AS Sampling errors for a binomial distribution (95% confidence level) 373 Table A6 Sampling ettors for a binomial distribution (99% confidence level) 373 Key equations 374 Glossary 379 Answers 383 Index 397