To understand how parental time investments influence child outcomes, we will first separate the contribution of other confounders such as:
- prenatal factors
- parental income and wealth
- parental employment
In some cases, individuals are linked across families and in networks. This allows us to compare:
- contributions from parental time inputs net of other factors
- school and neighbourhood characteristics.
These issues will be addressed through many state-of-the-art techniques that take care of selection and causation such as:
- treatment effects models
- siblings' fixed effects models
- instrumental variables
- propensity score matching.
To understand the factors at play we will:
- move forward statistical estimations of structural models of the technology of skills formation
- take advantage of the multi-level and longitudinal nature of the linked data to run sub-group analysis and heterogeneous effects - particularly across gender and class.