使用stata做多水平模型与纵向研究模型-经管之家官网！

[local]2[/local][local]2[/local]使用stata做多水平模型与纵向研究模型
Preliminaries
Review of linear regression
1.1Introduction........
1.2Isthere gender discrimination in faculty salaries?
1.3Independent-samples ttest.
1.4One-way analysis of variance
1.5Simplelinear regression
1.6Dummy variables....
1.7Multiple linear regression
1.8Interactions........
1.9Dummies for more than two groups.
1.10Other types of interactions ..... .
1.10.1Interaction between dummy variables.
1.10.2Interaction between continuous covariates
1.11Nonlinear effects..
1.12Residual diagnostics
1.13Summary and furtherreading
1.14Exercises ..........
IITwo-level linear models
2Variance-components models
xxi
xxv
xxxi
1
3
3
3
4
8
11
18
20
26
29
34
34
35
36
38
39
40
49
51 Contentsix
3Random-intercept models with covariates
3.1Introduction. . . . . . . . . . . . . . . . .
3.2Does smoking during pregnancy affect birthweight?
3.3The linear random-intercept model with covariates
3.3.1
3.3.2
Model specification. . . . . . . . . . . . .
Residual variance and intraclass correlation
3.4Estimation using Stata.
3.4.1
3.4.2
3.4.3
Using xtreg..
Using xtmixed.
Using gllamm.
3.5Coefficients of determination or variance explained
3.6Hypothesis tests and confidenceintervals.. . . . .
3.6.1Hypothesis tests forregression coefficients
91
91
91
94
94
96
97
97
99
100
102
104
104
Hypothesis tests forindividual regression coefficients105
Joint hypothesistests forseveral regression coefficients105
3.6.2
3.6.3
Predicted means and confidenceintervals. .
Hypothesis test forbetween-cluster variance
3.7Between and within effects....
Between-mother effects.
Within-mother effects ..
Relationsamong estimators
107
108
109
109
111
113
3.7.1
3.7.2
3.7.3
3.7.4
3.7.5
Endogeneity and different within- and between-mother effects114
Hausman endogeneity test.
3.8Fixed versusrandom effectsrevisited
3.9Residual diagnostics.........
3.10More on statistical inference forregression coefficients
122
124
125
129
3.10.1Consequences of using ordinary regression forclustered data129
3.10.2.:. Powerand sample-size determination
3.11Summary and furtherreading
3.12Exercises ............
130
132
132 viiiContents
2.1Introduction. . . . . . . . . . . . . . . . . . . . . . . .51
2.2How reliable are peak-expiratory-flow measurements?.52
2.3The variance-components model.. . . . . . . .54
2.3.1Model specification and path diagram54
,(
2.3.2Error components, variance components,and reliability.58
2.3.3Intraclass correlation58
2.4Fixed versusrandom effects61
2.5Estimation using Stata. .62
2.5.1Data preparation62
2.5.2Using xtreg..63
2.5.3Using xtmixed.65
2.5.4Using gllamm66
2.6Hypothesistests and confidence intervals.68
2.6.1Hypothesis test and confidence interval for the population
mean.............................68
2.6.2Hypothesistest and confidenceinterval forthe between-
cluster variance...69
2.7More on statistical inference. . . . . . .71
2.7.1.:. Different estimation methods71
2.7.2Inference forf3.........72
Estimate and standard error:Balanced case72
Estimate:Unbalanced case.74
2.8Crossed versus nested effects.. . . .75
2.9Assigning values to the random intercepts77
2.9.1Maximum likelihood estimation78
Implementation via OLSregression78
Implementation via the mean total residual79
2.9.2Empirical Bayes prediction..80
2.9.3.:. Empirical Bayes variances.83
2.10Summary and furtherreading85
2.11Exercises ............86 Contents
5.3.1
5.3.2
Missing data.. . . . .,. . . . . . . . . .
Time-varying and time-constant variables
5.4Time scalesin longitudinal data.. . .
5.5Random- and fixed-effectsapproaches
5.5.1
5.5.2
5.5.3
5.5.4
5.5.5
Correlated residuals..
Fixed-intercept model
Using xtreg
Using anova
Random-intercept model
Random-coefficient model
Marginal mean and covariance structure induced by ran-
dom effects.........................
Marginal mean and covariance structurefor random-intercept
xi
181
181
182
185
185
186
186
189
192
194
196
models.. . . . . . . . . . . . . . . . . . . . . ."196
Marginal mean and covariance structurefor random-coefficient
models.198
5.6Marginal modeling.......
5.6.1
5.6.2
Covariance structures.
Compound symmetric or exchangeable structure
Random-coefficient structure.. .
Autoregressive residual structure
Unstructured covariance matrix
Marginal modeling using Stata
5.7Autoregressive- or lagged-response models
5.8Hybrid approaches. . . . . . . . . . . . .
5.8.1
5.8.2
5.8.3
Autoregressive response and random effects
.:.Autoregressive responses and autoregressive residuals.
Autoregressive residuals and random or fixedeffects
199
199
200
201
201
201
202
204
206
206
206
207
5.9Missing data...........................207
5.9.1.:. Maximum likelihood estimation under MAR: A simulation207
5.10How dochildren grow?........................ "210 x
4
5
Random-coefficient models
4.1Introduction. . . . . . .
4.2How effective are differentschools?
4.3Separate linear regressionsforeach school
"
Contents
141
141
141
142
4.4Specification and interpretation of arandom-coefficient model146
4.4.1Specification of random-coefficient model. . . . . . .146
4.4.2Interpretation ofthe random-effects variances and covariances150
4.5EstiI;nation using Stata.
4.5.1
4.5.2
Using xtmixed.
Random-intercept model
Random-coefficient model
Using gllamm. . . . . .
Random-intercept model
Random-coefficient model
4.6Testing the slope variance
4.7Interpretation of estimates.
4.8Assigning valuesto the random intercepts and slopes
4.8.1
4.8.2
4.8.3
4.8.4
4.8.5
Maximum likelihood estimation
Empirical Bayes prediction.
Model visualization.
Residual diagnostics
Inferencesforindividual schools
4.9Two-stage model formulation..... .
4.10Some warnings about random-coefficient models.
4.11Summary and furtherreading
4.12Exercises ............
Longitudinal,panel,and growth-curve models
5.1Introduction.................
5.2How and why do wageschange over time?
5.3Data structure...............
153
153
153
155
157
157
157
159
159
161
161
162
164
165
167
168
171
173
173
179
179
180
181 Contentsxiii
6.10.1Conditional or residual intraclass correlation of the latent
responses ......256
6.10.2Median odds ratio.. .257
6.11Maximum likelihood estimation258
6.11.1.:.Adaptive quadrature258
6.11.2Some speed considerations261