1980
DOI: 10.4135/9781412984843
|View full text |Cite
|
Sign up to set email alerts
|

Log-Linear Models

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
222
0
14

Year Published

1988
1988
2015
2015

Publication Types

Select...
10

Relationship

0
10

Authors

Journals

citations
Cited by 432 publications
(237 citation statements)
references
References 0 publications
1
222
0
14
Order By: Relevance
“…We used log-linear models (Knoke & Burke, 1990) to test whether the observed frequency for the eight possible combinations presented above followed the expected pattern under the hypothesis of no association of nutritional status among household members. In case of rejection of the hypothesis of independence, that is, in case the observed distribution signi®-cantly differed from the expected one, a log-linear model was ®tted in order to incorporate the parameters of the associations father-child, mother-child and fathermother.…”
Section: Methodsmentioning
confidence: 99%
“…We used log-linear models (Knoke & Burke, 1990) to test whether the observed frequency for the eight possible combinations presented above followed the expected pattern under the hypothesis of no association of nutritional status among household members. In case of rejection of the hypothesis of independence, that is, in case the observed distribution signi®-cantly differed from the expected one, a log-linear model was ®tted in order to incorporate the parameters of the associations father-child, mother-child and fathermother.…”
Section: Methodsmentioning
confidence: 99%
“…This model is analogous to the saturated log-linear model described in texts on categorical data analysis (e.g., Agresti 1996;Bishop, Fienberg and Holland 1975;Knoke and Burke 1980).…”
Section: Describing the Age And Spatial Structures Of Migrationmentioning
confidence: 99%
“…The modelling process deletes higher order interaction terms until it finds the smallest number of interactions between the four variables that provide at least as good a model (the parsimonious model) as the saturated model. In addition, the results reveal any combination of autonomy and use of networks variables that are significantly related to increased use of domestic sourcing in any particular model (Knoke and Burke, 1980).…”
Section: Methodsmentioning
confidence: 90%