Thanks for these comments Kat! So your study would use naturally occurring variability in parenting behaviors. This makes a lot of sense and may be the best feasible approach, but note that naturally occurring variations may reflect other factors that would also influence the child's well-being, e.g., poverty, other children, parent's social status, marital stability, etc.
Maria