Innovative statistical methods for case-family design
The statistical analysis of data from families is challenging, due to the correlation of outcomes within families and the ascertainment of families. Typically outcome dependent sampling designs are used to increase the information on relationships between outcomes and genetic and environmental factors. The reasoning is in principle the same as for case-control designs. Indeed for recruitment of families to study family violence and social anxiety outcome dependent sampling design is applied.
With the social anxiety family project we have regular meetings to discuss the statistical analyses of their data, i.e. types of methods to be used and which methodological extensions are needed to answer relevant questions based on the family data.