Categorical

Categorical data poses many challenges in data analysis. One can think of categorical data as arising from continuous data, however instead of seeing the continuous data value, we can see only whether it is above or below certain values. So instead of modelling observed values we must model the probability of seeing a response in a given category. This requires fundamentally different mathematical models to continuous data.

Wright Dose Ltd can provide

Integration of categorical data scales with standard PKPD models, include random effects for variable patient sensitivity
Time series models for changes in the underlying data-generating process
Poisson-type models for count data, including random effects, overdispersion and models for modified zero incidence
Multivariate models for integrating multiple categorical measures such as different adverse events
Customized marginal distributions for unusual data arising from a mixture of processes

This site was last updated 21-Jun-2006