Log-Linear ModelsIntroduces methods for quantitative assessment of relationships among categoric variables in multivariable crosstabulations. Procedures to estimate and interpret effect parameters for hierarchical models are described for both the general loglinear model and its logit version. |
Contents
Editors Introduction | 5 |
The LogLinear Model | 11 |
Testing for Fit | 30 |
Applications to Substantive Problems | 42 |
Special Techniques with LogLinear Models | 63 |
Conclusions | 76 |
Other editions - View all
Common terms and phrases
algorithm analysis analyze baseline model calculated categoric variables Catholic causal model chi-square cohort effects College comparing conditional odds contingency table data in Table degrees of freedom dependent variable df's difference effect parameters equal Equation example expected cell frequencies expected frequencies expected odds ratio F₁ Fij's fit the data Fitted Marginals four-variable four-way table Goodman hierarchical models High School Graduate homicide independent iterative proportional fitting Knoke Less than High linear log-linear models logit model marginal distributions marginal homogeneity marginal table Markov chain membership mobility model 28 Models Fitted Non-Catholic nonsaturated models notation observed frequencies odds on voting panel Party Identification period polytomous variables procedure quasi-symmetry race relationship row and column sample saturated model significant social starting values statistical structural zeros substantive subtable symmetry Table 18 tau parameters three-way two-variable two-way table Type I error VM VM voluntary association voting turnout YPCJO