%PDF-1.5 % 4 0 obj << /S /GoTo /D (title.0) >> endobj 7 0 obj (Title Page) endobj 8 0 obj << /S /GoTo /D (copyright.0) >> endobj 11 0 obj (Copyright Notice) endobj 12 0 obj << /S /GoTo /D (dedication.0) >> endobj 15 0 obj (Dedication) endobj 16 0 obj << /S /GoTo /D (chapter*.5) >> endobj 19 0 obj (Table of Contents) endobj 20 0 obj << /S /GoTo /D (chapter*.6) >> endobj 23 0 obj (Introduction) endobj 24 0 obj << /S /GoTo /D (section*.7) >> endobj 27 0 obj (About the Centre for Multilevel Modelling) endobj 28 0 obj << /S /GoTo /D (section*.8) >> endobj 31 0 obj (Installing the MLwiN software) endobj 32 0 obj << /S /GoTo /D (section*.9) >> endobj 35 0 obj (MLwiN overview) endobj 36 0 obj << /S /GoTo /D (section*.10) >> endobj 39 0 obj (Enhancements in Version 2.26) endobj 40 0 obj << /S /GoTo /D (section*.11) >> endobj 43 0 obj (Estimation) endobj 44 0 obj << /S /GoTo /D (section*.12) >> endobj 47 0 obj (Exploring, importing and exporting data) endobj 48 0 obj << /S /GoTo /D (section*.13) >> endobj 51 0 obj (Improved ease of use) endobj 52 0 obj << /S /GoTo /D (section*.14) >> endobj 55 0 obj (MLwiN Help) endobj 56 0 obj << /S /GoTo /D (section*.15) >> endobj 59 0 obj (Compatibility with existing MLn software) endobj 60 0 obj << /S /GoTo /D (section*.16) >> endobj 63 0 obj (Macros) endobj 64 0 obj << /S /GoTo /D (section*.17) >> endobj 67 0 obj (The structure of the User's Guide) endobj 68 0 obj << /S /GoTo /D (section*.18) >> endobj 71 0 obj (Acknowledgements) endobj 72 0 obj << /S /GoTo /D (section*.19) >> endobj 75 0 obj (Further information about multilevel modelling) endobj 76 0 obj << /S /GoTo /D (section*.20) >> endobj 79 0 obj (Technical Support) endobj 80 0 obj << /S /GoTo /D (chapter.1) >> endobj 83 0 obj (Introducing Multilevel Models) endobj 84 0 obj << /S /GoTo /D (section.1.1) >> endobj 87 0 obj (Multilevel data structures) endobj 88 0 obj << /S /GoTo /D (section.1.2) >> endobj 91 0 obj (Consequences of ignoring a multilevel structure) endobj 92 0 obj << /S /GoTo /D (section.1.3) >> endobj 95 0 obj (Levels of a data structure) endobj 96 0 obj << /S /GoTo /D (section.1.4) >> endobj 99 0 obj (An introductory description of multilevel modelling) endobj 100 0 obj << /S /GoTo /D (chapter.2) >> endobj 103 0 obj (Introduction to Multilevel Modelling) endobj 104 0 obj << /S /GoTo /D (section.2.1) >> endobj 107 0 obj (The tutorial data set) endobj 108 0 obj << /S /GoTo /D (section.2.2) >> endobj 111 0 obj (Opening the worksheet and looking at the data) endobj 112 0 obj << /S /GoTo /D (section.2.3) >> endobj 115 0 obj (Comparing two groups) endobj 116 0 obj << /S /GoTo /D (section.2.4) >> endobj 119 0 obj (Comparing more than two groups: Fixed effects models) endobj 120 0 obj << /S /GoTo /D (section.2.5) >> endobj 123 0 obj (Comparing means: Random effects or multilevel model) endobj 124 0 obj << /S /GoTo /D (section*.22) >> endobj 127 0 obj (Chapter learning outcomes) endobj 128 0 obj << /S /GoTo /D (chapter.3) >> endobj 131 0 obj (Residuals) endobj 132 0 obj << /S /GoTo /D (section.3.1) >> endobj 135 0 obj (What are multilevel residuals?) endobj 136 0 obj << /S /GoTo /D (section.3.2) >> endobj 139 0 obj (Calculating residuals in MLwiN) endobj 140 0 obj << /S /GoTo /D (section.3.3) >> endobj 143 0 obj (Normal plots) endobj 144 0 obj << /S /GoTo /D (section*.23) >> endobj 147 0 obj (Chapter learning outcomes) endobj 148 0 obj << /S /GoTo /D (chapter.4) >> endobj 151 0 obj (Random Intercept and Random Slope Models) endobj 152 0 obj << /S /GoTo /D (section.4.1) >> endobj 155 0 obj (Random intercept models) endobj 156 0 obj << /S /GoTo /D (section.4.2) >> endobj 159 0 obj (Graphing predicted school lines from a random intercept model) endobj 160 0 obj << /S /GoTo /D (section.4.3) >> endobj 163 0 obj (The effect of clustering on the standard errors of coefficients) endobj 164 0 obj << /S /GoTo /D (section.4.4) >> endobj 167 0 obj (Does the coefficient of standlrt vary across schools? Introducing a random slope) endobj 168 0 obj << /S /GoTo /D (section.4.5) >> endobj 171 0 obj (Graphing predicted school lines from a random slope model) endobj 172 0 obj << /S /GoTo /D (section*.24) >> endobj 175 0 obj (Chapter learning outcomes) endobj 176 0 obj << /S /GoTo /D (chapter.5) >> endobj 179 0 obj (Graphical Procedures for Exploring the Model) endobj 180 0 obj << /S /GoTo /D (section.5.1) >> endobj 183 0 obj (Displaying multiple graphs) endobj 184 0 obj << /S /GoTo /D (section.5.2) >> endobj 187 0 obj (Highlighting in graphs) endobj 188 0 obj << /S /GoTo /D (section*.25) >> endobj 191 0 obj (Chapter learning outcomes) endobj 192 0 obj << /S /GoTo /D (chapter.6) >> endobj 195 0 obj (Contextual Effects) endobj 196 0 obj << /S /GoTo /D (section.6.1) >> endobj 199 0 obj (The impact of school gender on girls' achievement) endobj 200 0 obj << /S /GoTo /D (section.6.2) >> endobj 203 0 obj (Contextual effects of school intake ability averages) endobj 204 0 obj << /S /GoTo /D (section*.26) >> endobj 207 0 obj (Chapter learning outcomes) endobj 208 0 obj << /S /GoTo /D (chapter.7) >> endobj 211 0 obj (Modelling the Variance as a Function of Explanatory Variables) endobj 212 0 obj << /S /GoTo /D (section.7.1) >> endobj 215 0 obj (A level 1 variance function for two groups) endobj 216 0 obj << /S /GoTo /D (section.7.2) >> endobj 219 0 obj (Variance functions at level 2) endobj 220 0 obj << /S /GoTo /D (section.7.3) >> endobj 223 0 obj (Further elaborating the model for the student-level variance) endobj 224 0 obj << /S /GoTo /D (section*.27) >> endobj 227 0 obj (Chapter learning outcomes) endobj 228 0 obj << /S /GoTo /D (chapter.8) >> endobj 231 0 obj (Getting Started with your Data) endobj 232 0 obj << /S /GoTo /D (section.8.1) >> endobj 235 0 obj (Inputting your data set into MLwiN) endobj 236 0 obj << /S /GoTo /D (section*.28) >> endobj 239 0 obj (Reading in an ASCII text data file) endobj 240 0 obj << /S /GoTo /D (section*.29) >> endobj 243 0 obj (Common problems that can occur in reading ASCII data from a text file) endobj 244 0 obj << /S /GoTo /D (section*.30) >> endobj 247 0 obj (Pasting data into a worksheet from the clipboard) endobj 248 0 obj << /S /GoTo /D (section*.31) >> endobj 251 0 obj (Naming columns) endobj 252 0 obj << /S /GoTo /D (section*.32) >> endobj 255 0 obj (Adding category names) endobj 256 0 obj << /S /GoTo /D (section*.33) >> endobj 259 0 obj (Missing data) endobj 260 0 obj << /S /GoTo /D (section*.34) >> endobj 263 0 obj (Unit identification columns) endobj 264 0 obj << /S /GoTo /D (section*.35) >> endobj 267 0 obj (Saving the worksheet) endobj 268 0 obj << /S /GoTo /D (section*.36) >> endobj 271 0 obj (Sorting your data set) endobj 272 0 obj << /S /GoTo /D (section.8.2) >> endobj 275 0 obj (Fitting models in MLwiN) endobj 276 0 obj << /S /GoTo /D (section*.37) >> endobj 279 0 obj (What are you trying to model?) endobj 280 0 obj << /S /GoTo /D (section*.38) >> endobj 283 0 obj (Do you really need to fit a multilevel model?) endobj 284 0 obj << /S /GoTo /D (section*.39) >> endobj 287 0 obj (Have you built up your model from a variance components model?) endobj 288 0 obj << /S /GoTo /D (section*.40) >> endobj 291 0 obj (Have you centred your predictor variables?) endobj 292 0 obj << /S /GoTo /D (section*.41) >> endobj 295 0 obj (Chapter learning outcomes) endobj 296 0 obj << /S /GoTo /D (chapter.9) >> endobj 299 0 obj (Logistic Models for Binary and Binomial Responses) endobj 300 0 obj << /S /GoTo /D (section.9.1) >> endobj 303 0 obj (Introduction and description of the example data) endobj 304 0 obj << /S /GoTo /D (section.9.2) >> endobj 307 0 obj (Single-level logistic regression) endobj 308 0 obj << /S /GoTo /D (section*.42) >> endobj 311 0 obj (Link functions) endobj 312 0 obj << /S /GoTo /D (section*.43) >> endobj 315 0 obj (Interpretation of coefficients) endobj 316 0 obj << /S /GoTo /D (section*.44) >> endobj 319 0 obj (Fitting a single-level logit model in MLwiN) endobj 320 0 obj << /S /GoTo /D (section*.45) >> endobj 323 0 obj (A probit model) endobj 324 0 obj << /S /GoTo /D (section.9.3) >> endobj 327 0 obj (A two-level random intercept model) endobj 328 0 obj << /S /GoTo /D (section*.46) >> endobj 331 0 obj (Model specification) endobj 332 0 obj << /S /GoTo /D (section*.47) >> endobj 335 0 obj (Estimation procedures) endobj 336 0 obj << /S /GoTo /D (section*.48) >> endobj 339 0 obj (Fitting a two-level random intercept model in MLwiN) endobj 340 0 obj << /S /GoTo /D (section*.49) >> endobj 343 0 obj (Variance partition coefficient) endobj 344 0 obj << /S /GoTo /D (section*.50) >> endobj 347 0 obj (Adding further explanatory variables) endobj 348 0 obj << /S /GoTo /D (section.9.4) >> endobj 351 0 obj (A two-level random coefficient model) endobj 352 0 obj << /S /GoTo /D (section.9.5) >> endobj 355 0 obj (Modelling binomial data) endobj 356 0 obj << /S /GoTo /D (section*.51) >> endobj 359 0 obj (Modelling district-level variation with district-level proportions) endobj 360 0 obj << /S /GoTo /D (section*.52) >> endobj 363 0 obj (Creating a district-level data set) endobj 364 0 obj << /S /GoTo /D (section*.53) >> endobj 367 0 obj (Fitting the model) endobj 368 0 obj << /S /GoTo /D (section*.54) >> endobj 371 0 obj (Chapter learning outcomes) endobj 372 0 obj << /S /GoTo /D (chapter.10) >> endobj 375 0 obj (Multinomial Logistic Models for Unordered Categorical Responses) endobj 376 0 obj << /S /GoTo /D (section.10.1) >> endobj 379 0 obj (Introduction) endobj 380 0 obj << /S /GoTo /D (section.10.2) >> endobj 383 0 obj (Single-level multinomial logistic regression) endobj 384 0 obj << /S /GoTo /D (section.10.3) >> endobj 387 0 obj (Fitting a single-level multinomial logistic model in MLwiN) endobj 388 0 obj << /S /GoTo /D (section.10.4) >> endobj 391 0 obj (A two-level random intercept multinomial logistic regression model) endobj 392 0 obj << /S /GoTo /D (section.10.5) >> endobj 395 0 obj (Fitting a two-level random intercept model) endobj 396 0 obj << /S /GoTo /D (section*.55) >> endobj 399 0 obj (Chapter learning outcomes) endobj 400 0 obj << /S /GoTo /D (chapter.11) >> endobj 403 0 obj (Fitting an Ordered Category Response Model) endobj 404 0 obj << /S /GoTo /D (section.11.1) >> endobj 407 0 obj (Introduction) endobj 408 0 obj << /S /GoTo /D (section.11.2) >> endobj 411 0 obj (An analysis using the traditional approach) endobj 412 0 obj << /S /GoTo /D (section.11.3) >> endobj 415 0 obj (A single-level model with an ordered categorical response variable) endobj 416 0 obj << /S /GoTo /D (section.11.4) >> endobj 419 0 obj (A two-level model) endobj 420 0 obj << /S /GoTo /D (section*.57) >> endobj 423 0 obj (Chapter learning outcomes) endobj 424 0 obj << /S /GoTo /D (chapter.12) >> endobj 427 0 obj (Modelling Count Data) endobj 428 0 obj << /S /GoTo /D (section.12.1) >> endobj 431 0 obj (Introduction) endobj 432 0 obj << /S /GoTo /D (section.12.2) >> endobj 435 0 obj (Fitting a simple Poisson model) endobj 436 0 obj << /S /GoTo /D (section.12.3) >> endobj 439 0 obj (A three-level analysis) endobj 440 0 obj << /S /GoTo /D (section.12.4) >> endobj 443 0 obj (A two-level model using separate country terms) endobj 444 0 obj << /S /GoTo /D (section.12.5) >> endobj 447 0 obj (Some issues and problems for discrete response models) endobj 448 0 obj << /S /GoTo /D (section*.59) >> endobj 451 0 obj (Chapter learning outcomes) endobj 452 0 obj << /S /GoTo /D (chapter.13) >> endobj 455 0 obj (Fitting Models to Repeated Measures Data) endobj 456 0 obj << /S /GoTo /D (section.13.1) >> endobj 459 0 obj (Introduction) endobj 460 0 obj << /S /GoTo /D (section.13.2) >> endobj 463 0 obj (A basic model) endobj 464 0 obj << /S /GoTo /D (section.13.3) >> endobj 467 0 obj (A linear growth curve model) endobj 468 0 obj << /S /GoTo /D (section.13.4) >> endobj 471 0 obj (Complex level 1 variation) endobj 472 0 obj << /S /GoTo /D (section.13.5) >> endobj 475 0 obj (Repeated measures modelling of non-linear polynomial growth) endobj 476 0 obj << /S /GoTo /D (section*.66) >> endobj 479 0 obj (Chapter learning outcomes) endobj 480 0 obj << /S /GoTo /D (chapter.14) >> endobj 483 0 obj (Multivariate Response Models) endobj 484 0 obj << /S /GoTo /D (section.14.1) >> endobj 487 0 obj (Introduction) endobj 488 0 obj << /S /GoTo /D (section.14.2) >> endobj 491 0 obj (Specifying a multivariate model) endobj 492 0 obj << /S /GoTo /D (section.14.3) >> endobj 495 0 obj (Setting up the basic model) endobj 496 0 obj << /S /GoTo /D (section.14.4) >> endobj 499 0 obj (A more elaborate model) endobj 500 0 obj << /S /GoTo /D (section.14.5) >> endobj 503 0 obj (Multivariate models for discrete responses) endobj 504 0 obj << /S /GoTo /D (section*.68) >> endobj 507 0 obj (Chapter learning outcomes) endobj 508 0 obj << /S /GoTo /D (chapter.15) >> endobj 511 0 obj (Diagnostics for Multilevel Models) endobj 512 0 obj << /S /GoTo /D (section.15.1) >> endobj 515 0 obj (Introduction) endobj 516 0 obj << /S /GoTo /D (section.15.2) >> endobj 519 0 obj (Diagnostics plotting: Deletion residuals, influence and leverage) endobj 520 0 obj << /S /GoTo /D (section.15.3) >> endobj 523 0 obj (A general approach to data exploration) endobj 524 0 obj << /S /GoTo /D (section*.70) >> endobj 527 0 obj (Chapter learning outcomes) endobj 528 0 obj << /S /GoTo /D (chapter.16) >> endobj 531 0 obj (An Introduction to Simulation Methods of Estimation) endobj 532 0 obj << /S /GoTo /D (section.16.1) >> endobj 535 0 obj (An illustration of parameter estimation with Normally distributed data) endobj 536 0 obj << /S /GoTo /D (section.16.2) >> endobj 539 0 obj (Generating random numbers in MLwiN) endobj 540 0 obj << /S /GoTo /D (section*.73) >> endobj 543 0 obj (Chapter learning outcomes) endobj 544 0 obj << /S /GoTo /D (chapter.17) >> endobj 547 0 obj (Bootstrap Estimation) endobj 548 0 obj << /S /GoTo /D (section.17.1) >> endobj 551 0 obj (Introduction) endobj 552 0 obj << /S /GoTo /D (section.17.2) >> endobj 555 0 obj (Understanding the iterated bootstrap) endobj 556 0 obj << /S /GoTo /D (section.17.3) >> endobj 559 0 obj (An example of bootstrapping using MLwiN) endobj 560 0 obj << /S /GoTo /D (section.17.4) >> endobj 563 0 obj (Diagnostics and confidence intervals) endobj 564 0 obj << /S /GoTo /D (section.17.5) >> endobj 567 0 obj (Nonparametric bootstrapping) endobj 568 0 obj << /S /GoTo /D (section*.76) >> endobj 571 0 obj (Chapter learning outcomes) endobj 572 0 obj << /S /GoTo /D (chapter.18) >> endobj 575 0 obj (Modelling Cross-classified Data) endobj 576 0 obj << /S /GoTo /D (section.18.1) >> endobj 579 0 obj (An introduction to cross-classification) endobj 580 0 obj << /S /GoTo /D (section.18.2) >> endobj 583 0 obj (How cross-classified models are implemented in MLwiN) endobj 584 0 obj << /S /GoTo /D (section.18.3) >> endobj 587 0 obj (Some computational considerations) endobj 588 0 obj << /S /GoTo /D (section.18.4) >> endobj 591 0 obj (Modelling a two-way classification: An example) endobj 592 0 obj << /S /GoTo /D (section.18.5) >> endobj 595 0 obj (Other aspects of the SETX command) endobj 596 0 obj << /S /GoTo /D (section.18.6) >> endobj 599 0 obj (Reducing storage overhead by grouping) endobj 600 0 obj << /S /GoTo /D (section.18.7) >> endobj 603 0 obj (Modelling a multi-way cross-classification) endobj 604 0 obj << /S /GoTo /D (section.18.8) >> endobj 607 0 obj (MLwiN commands for cross-classifications) endobj 608 0 obj << /S /GoTo /D (section*.80) >> endobj 611 0 obj (Chapter learning outcomes) endobj 612 0 obj << /S /GoTo /D (chapter.19) >> endobj 615 0 obj (Multiple Membership Models) endobj 616 0 obj << /S /GoTo /D (section.19.1) >> endobj 619 0 obj (A simple multiple membership model) endobj 620 0 obj << /S /GoTo /D (section.19.2) >> endobj 623 0 obj (MLwiN commands for multiple membership models) endobj 624 0 obj << /S /GoTo /D (section*.81) >> endobj 627 0 obj (Chapter learning outcomes) endobj 628 0 obj << /S /GoTo /D (chapter.20) >> endobj 631 0 obj (Bibliography) endobj 632 0 obj << /S /GoTo /D (chapter.21) >> endobj 635 0 obj (Index) endobj 636 0 obj << /S /GoTo /D [637 0 R /Fit] >> endobj 640 0 obj << /Length 484 /Filter /FlateDecode >> stream xڅRMs@WTF# Сl8tNJ{ObO\saA(rXNyޡ[^fՐfr:9^}xh>V?֗uq[p`3)#~ԐΡVz!׳ ܢZ%AktV5YB#nJLO89zE[8FXpmyTy i~W/cm8M|5-7[ӶM}0eC7vû:ާz/v;$CfU*a1&Xa]sruMS7s~p6n_2/a:iJ7]G屢.C-c3}; [G&x.|>c"&6 1dz6y Y)Oulǧѓ\|6;b%Is2{>s4tL:<"ZxMOzϽ0 endstream endobj 637 0 obj << /Type /Page /Contents 640 0 R /Resources 639 0 R /MediaBox [0 0 595.276 841.89] /Parent 646 0 R >> endobj 638 0 obj << /Type /XObject /Subtype /Image /Width 300 /Height 300 /BitsPerComponent 8 /ColorSpace [/Indexed /DeviceRGB 255 647 0 R] /Length 5313 /Filter /FlateDecode >> stream x{\Sg,i:Z(QMQlQԆKJ GȵHiVM/SˌcM/[GkvmUvٶ{Br I$<{mK7l0B}7_߹OnO9ܜ+3R!nnj i%fUlm Vn\*oܗ6&Silrs>ۦ-NrT1U[YcnCbwX:nm QM_hc]|*_'Z")mUkL l6U\_g&V[U1 m\gAnGl [vm^D7x@SiPU]ػF+ߓNI_ݸ*]uD1Zޤ_cmQ 9= sM5oO0dܞ;7RVm$(ۧ_*؝-4qЁ!Xz~cGG/Sϰ:9&9mj519ҵUTe4U:ʪ;K#naX%rz@ǬjȔ@tj|FITBi(_#'FrUڢ)NϢjsкFu`ꚦl]W 'ԪAjKkN+P&TSVjót0O~?^tvEJsJUEF6N(j|LB~kQQ7]-RjJMr///i*)0I1Ei ;:s%j~5Jr;TrCoU=5RJڮv%hRAb,H߄ IL̄`VggTt#Gt!uTJ*նVUm[}0aȉi3H3oClcZmcqHnwV"C\Nk+7#\Vq
ԣ=UiCR8U&}:{#(lU?ILd
5ͅmCvL3{GqδZUQ'.e
GqQmo Vꉰ[ 7V؊S:jnzÑV.dZ'REr]5i+>+t5[X"ɸT