Davina Hill
2012-Jun-03 15:03 UTC
[R] Multiple imputation, multinomial response & random effects
Dear R-group, Could somebody recommend a package that can deal with a multinomial response variable (choice of breeding tactic in mice, which has four unordered levels), multiply-imputed data (generated using the Amelia package) and two non-nested random effects: individual identity (133 individuals made up to four choices each) and year (for which there are six levels and sample size varies between years)? I've tried sabreR, drm, mixcat and mlogit but none of them seem able to accommodate multiply-imputed datasets. The most promising package I’ve found so far is Zelig, which can handle multiply-imputed data and either multinomial responses OR random effects (but seemingly not both). I could randomly select one case per individual and run the following model (with year as a fixed rather than random effect): z.out <- zelig(as.factor(tactic) ~ mass+age+year, model = "mlogit", data = a.out$imputations) but it probably isn't the most elegant solution. Does anybody have any suggestions? Thanks in advance, Davina ------------------------------------------ Animal, Plant and Environmental Sciences University of the Witwatersrand South Africa <html><p><font face = "verdana" size = "0.8" color = "navy">This communication is intended for the addressee only. It is confidential. If you have received this communication in error, please notify us immediately and destroy the original message. You may not copy or disseminate this communication without the permission of the University. Only authorized signatories are competent to enter into agreements on behalf of the University and recipients are thus advised that the content of this message may not be legally binding on the University and may contain the personal views and opinions of the author, which are not necessarily the views and opinions of The University of the Witwatersrand, Johannesburg. All agreements between the University and outsiders are subject to South African Law unless the University agrees in writing to the contrary.</font></p></html> [[alternative HTML version deleted]]
Ruben van eijk
2014-Jan-17 09:16 UTC
[R] Multiple imputation, multinomial response & random effects
Dear Davina, Unfortunately (or luckily), I have almost the exact same problem. I want to do a multilevel analysis with imputed data and both include mixed and random effects in the regression model. I have imputed my data with de Hmisc package (aregImpute), however, the rest of the functions does not support to analyse the data with random effects. I saw that your question is from 2012, but you still did not receive a public answer. Do you have one by now, or did you find a solution for this problem on your own? Best regards, Ruben Op zondag 3 juni 2012 17:03:21 UTC+2 schreef Davina Hill:> > Dear R-group, > > Could somebody recommend a package that can deal with a multinomial > response variable (choice of breeding tactic in mice, which has four > unordered levels), multiply-imputed data (generated using the Amelia > package) and two non-nested random effects: individual identity (133 > individuals made up to four choices each) and year (for which there are six > levels and sample size varies between years)? > > I've tried sabreR, drm, mixcat and mlogit but none of them seem able to > accommodate multiply-imputed datasets. The most promising package I?ve > found so far is Zelig, which can handle multiply-imputed data and either > multinomial responses OR random effects (but seemingly not both). I could > randomly select one case per individual and run the following model (with > year as a fixed rather than random effect): > > z.out <- zelig(as.factor(tactic) ~ mass+age+year, model = "mlogit", data = > a.out$imputations) > > but it probably isn't the most elegant solution. Does anybody have any > suggestions? > > Thanks in advance, > Davina > > ------------------------------------------ > Animal, Plant and Environmental Sciences > University of the Witwatersrand > South Africa > > <html><p><font face = "verdana" size = "0.8" color = "navy">This > communication is intended for the addressee only. It is confidential. If > you have received this communication in error, please notify us immediately > and destroy the original message. You may not copy or disseminate this > communication without the permission of the University. Only authorized > signatories are competent to enter into agreements on behalf of the > University and recipients are thus advised that the content of this message > may not be legally binding on the University and may contain the personal > views and opinions of the author, which are not necessarily the views and > opinions of The University of the Witwatersrand, Johannesburg. All > agreements between the University and outsiders are subject to South > African Law unless the University agrees in writing to the > contrary.</font></p></html> > > [[alternative HTML version deleted]] > >