Additionally, an overall measure of Medicare service use enabled us to control for differences in health service use more generally.
Measuring tree branches using "evolutionary time" allows us to control for variable rates of evolution across branches.
Secondly, we estimate the effects of parental illness using a panel dataset that allows us to control for unobserved heterogeneity at the household level.
Nevertheless, the use of the SCCS method allows us to control for time-independent confounders, making our estimations more robust.
Finally, the use of multi-level modeling allowed us to control for clustering within practices and to more clearly separate patient, provider and practice factors.
Panel regressions have the additional advantage of allowing us to control for investor-specific skills using fixed effects and overall market conditions using time effects.
For longitudinal predictions, this also enables us to control for initial levels of substance use or psychological distress.
For individual-level analysis, we used a multilevel logistic random-effects model which allowed us to control for clustering effects at the PSU and district levels.
This allowed us to control for mother plant effects along the distance gradient.
These questionnaires allowed us to control for personality traits and behavioral tendencies that are relevant in decision-making processes.
This allowed us to control for changes that would occur due to typical development and learning during the school year.
