For example, X has Y& Z, and further y& z have a,b,c & d,e,f respectively...and then definitely each a,b,c and d,e,f has its own questions. Reason to do like this is the fear that i may not have existance of variable X in my population, despite have B & C.
Based on what you have written I think the answer is no. Imputation is not suitable to recover 100% missing data. It is used to recover a few missing values.
There is a statistical procedure called Factor Analysis that generates factors. I am assuming that this is not relevant to this question.
Your question seems to be regarding the higher order constructs, which are formed by multi-dimensional constructs i.e., construct has multiple dimensions, which further have multiple questions. Read HOC in 'A primer on Structural Equation Modeling - Guide for SmartPLS.
For imputation follow the answer of Timothy.
Jazzak Allah Wali ur Rehman, sahib. Thank you so much. can i get your email id?
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning