Matches in SemOpenAlex for { <https://semopenalex.org/work/W137018658> ?p ?o ?g. }
Showing items 1 to 78 of
78
with 100 items per page.
- W137018658 abstract "Background : Logistic regression is one of the most widely used models to analyze the relation between one or more explanatory variables and a categorical response in the field of epidemiology, health and medicine. When there is strong correlation among explanatory variables, i.e.multicollinearity, the efficiency of model reduces considerably. The objective of this research was to employ latent variables to reduce the effect of multicollinearity in analysis of a case-control study about breast cancer risk factors. Methods : The data belonged to a case-control study in which 300 women with breast cancer were compared to same number of controls. To assess the effect of multicollinearity, five highly correlated quantitative variables were selected. Ordinary logistic regression with collinear data was compared to two models contain latent variables were generated using either factor analysis or principal components analysis. Estimated standard errors of parameters were selected to compare the efficiency of models. We also conducted a simulation study in order to compare the efficiency of models with and without latent factors. All analyses were carried out using S-plus. Results : Logistic regression based on five primary variables showed an unusual odds ratios for age at first pregnancy (OR=67960, 95%CI: 10184-453503) and for total length of breast feeding (OR=0). On the other hand the parameters estimated for logistic regression on latent variables generated by both factor analysis and principal components analysis were statistically significant (P<0.003). Their standard errors were smaller than that of ordinary logistic regression on original variables. The simulation showed that in the case of normal error and 58% reliability the logistic regression based on latent variables is more efficient than that model for collinear variables. Conclusions : This research indicated that logistic regression based on latent variables is more efficient than logistic regression based on original collinear variables." @default.
- W137018658 created "2016-06-24" @default.
- W137018658 creator A5006402635 @default.
- W137018658 creator A5020090871 @default.
- W137018658 creator A5058538599 @default.
- W137018658 creator A5060504871 @default.
- W137018658 date "2008-01-01" @default.
- W137018658 modified "2023-09-23" @default.
- W137018658 title "Using latent variables in logistic regression to reduce multicollinearity, A case-control example: breast cancer risk factors" @default.
- W137018658 doi "https://doi.org/10.2427/5857" @default.
- W137018658 hasPublicationYear "2008" @default.
- W137018658 type Work @default.
- W137018658 sameAs 137018658 @default.
- W137018658 citedByCount "3" @default.
- W137018658 countsByYear W1370186582014 @default.
- W137018658 countsByYear W1370186582019 @default.
- W137018658 crossrefType "journal-article" @default.
- W137018658 hasAuthorship W137018658A5006402635 @default.
- W137018658 hasAuthorship W137018658A5020090871 @default.
- W137018658 hasAuthorship W137018658A5058538599 @default.
- W137018658 hasAuthorship W137018658A5060504871 @default.
- W137018658 hasConcept C105795698 @default.
- W137018658 hasConcept C114494560 @default.
- W137018658 hasConcept C121608353 @default.
- W137018658 hasConcept C126322002 @default.
- W137018658 hasConcept C149782125 @default.
- W137018658 hasConcept C151956035 @default.
- W137018658 hasConcept C152732102 @default.
- W137018658 hasConcept C152877465 @default.
- W137018658 hasConcept C189285262 @default.
- W137018658 hasConcept C33923547 @default.
- W137018658 hasConcept C51167844 @default.
- W137018658 hasConcept C5274069 @default.
- W137018658 hasConcept C530470458 @default.
- W137018658 hasConcept C71924100 @default.
- W137018658 hasConceptScore W137018658C105795698 @default.
- W137018658 hasConceptScore W137018658C114494560 @default.
- W137018658 hasConceptScore W137018658C121608353 @default.
- W137018658 hasConceptScore W137018658C126322002 @default.
- W137018658 hasConceptScore W137018658C149782125 @default.
- W137018658 hasConceptScore W137018658C151956035 @default.
- W137018658 hasConceptScore W137018658C152732102 @default.
- W137018658 hasConceptScore W137018658C152877465 @default.
- W137018658 hasConceptScore W137018658C189285262 @default.
- W137018658 hasConceptScore W137018658C33923547 @default.
- W137018658 hasConceptScore W137018658C51167844 @default.
- W137018658 hasConceptScore W137018658C5274069 @default.
- W137018658 hasConceptScore W137018658C530470458 @default.
- W137018658 hasConceptScore W137018658C71924100 @default.
- W137018658 hasIssue "1" @default.
- W137018658 hasLocation W1370186581 @default.
- W137018658 hasOpenAccess W137018658 @default.
- W137018658 hasPrimaryLocation W1370186581 @default.
- W137018658 hasRelatedWork W1600904143 @default.
- W137018658 hasRelatedWork W1664440805 @default.
- W137018658 hasRelatedWork W1936198225 @default.
- W137018658 hasRelatedWork W2143570340 @default.
- W137018658 hasRelatedWork W2163913151 @default.
- W137018658 hasRelatedWork W2195347453 @default.
- W137018658 hasRelatedWork W2316630697 @default.
- W137018658 hasRelatedWork W2364190653 @default.
- W137018658 hasRelatedWork W2367214771 @default.
- W137018658 hasRelatedWork W2372170925 @default.
- W137018658 hasRelatedWork W2503713407 @default.
- W137018658 hasRelatedWork W282016921 @default.
- W137018658 hasRelatedWork W289214573 @default.
- W137018658 hasRelatedWork W3014509598 @default.
- W137018658 hasRelatedWork W3133529159 @default.
- W137018658 hasRelatedWork W3161269547 @default.
- W137018658 hasRelatedWork W614020064 @default.
- W137018658 hasRelatedWork W2596970929 @default.
- W137018658 hasRelatedWork W2739255552 @default.
- W137018658 hasRelatedWork W2774452861 @default.
- W137018658 hasVolume "5" @default.
- W137018658 isParatext "false" @default.
- W137018658 isRetracted "false" @default.
- W137018658 magId "137018658" @default.
- W137018658 workType "article" @default.