ABSTRACT
The model is for the number of days living at home prior to admission to an institution for long-term care (DAYS).
The model was fitted using tobit analysis. Chapter 2 explained that the use of tobit analysis was necessary because the distribution of the variable is censored. The variable is censored because the results could not otherwise be produced at this stage without imposing an early cut-off point. The cut-off point was the 22nd month, for the case that was in the median position in the distribution by date of referral.
