[R] Error when running duplicate scale imputation for multilevel data

Yvonne Yock yvonne.yock at gmail.com
Wed Feb 7 17:22:07 CET 2018


Hi,

I am working with a multiple-item questionnaire. I have previously done
item-level multiple imputation using MICE in R and right now I am
attempting duplicate-scale imputation based on the guidelines listed in
Enders's applied missing data analysis book.

I use MICE to do MI as it allows me to specify school effect as I am
working with multilevel data; my respondents come from different schools.

I had no problem running the item-level MI with the school effect. However,
I encounter problem running the duplicate-scale MI when I included the
school effect. I saw this error message '*Error in x + diag(diag(x) *
ridge) : non-conformable arrays*'. I did not get any error message when I
re-ran without school effect.

What could be the reason why I am facing difficulties running duplicate
scale imputation when I specify school effect?

Thanks. ( I have not included any codes as I am not sure if it will be
useful)
Regards,
Yvonne

	[[alternative HTML version deleted]]



More information about the R-help mailing list