Sedentary life in children is evaluated depending on television use and/ playing video games, of computer use and hours spent on homework.
How can we score it, do we need to put separately; High levels indicate television use (> 4 hours a day), computer use (> 3 hours a day), hours spent on homework (> 3 hours a day)
OR
Weighted average time of the 3 variables
OR
Average Screen time (television use and computer use) and Non Screen time (others like homework hours)
I recomend use observational metodology. You can define some categories related with (tv use, video games, and computer use). Then you can measure time allocated to each category, then you can calculate relative time allocated.
Whenever dealing with indicators like this, which are not necessarily part of a "scale", I recommend using latent class or clustering methods to create distinct groups of individuals. In your case, I suspect one such group will correspond to a sedentary lifestyle (low rates of occurrence across all items). The key to this type of analysis is to have a decent sample size, but with only a few indicators about 100 should be sufficient. SPSS can do Finite Mixture Models (Two-Step Clustering), SAS, MPlus and R all can do Latent Class Analysis. If longitudinal data is involved, you can do Latent Transition Analysis, which is an extention of the basic LCA model.
My sample sizes in research on hemophilic children were very small but I compensated with two classmates for every child: Article School Absences and School Achievements in Children with Con...
parents can fill in questionnaires too.