# markmfredrickson/natural-experiments-in-genetics

moving things around in presentation

 @@ -46,11 +46,22 @@ $n =$ \end{itemize} \end{frame} +\begin{frame}{ACE Model} + ACE summarizes the difference with the difference of standardized + variables (i.e. correlation) \citep{Rodgers:1988fk} for monozygotic ($\rmz$) + and dizygotic ($\rdz$) twin pairs: + \begin{description} + \item[Genetic] $A = 2 (\rmz - \rdz)$ + \item[Environment] $C = 2 \rdz - \rmz$ + \item[Error] $E = 1 - \rmz$ + \end{description} + % CHECK THIS: By implication, $A^2 + C^2 + E^2 = 1$. + By design $A + C + E = 1$. Variations fit using SEM and Bayesian methods. +\end{frame} + \begin{frame}[fragile]{Fitting the ACE Model} \begin{itemize} - \item ACE summarizes the difference with the difference of standardized - variables (i.e. correlation) \citep{Rodgers:1988fk} \item ACE estimates with bootstrapped standard errors: <>= library(boot) # need this to get formatted errors @@ -63,3 +74,26 @@ cat(tmp.str) \end{itemize} \end{frame} + +\begin{frame}{ACE Assumptions I} + \begin{description} + \item[Randomized] Twin pairs are randomly assigned to be MZ or DZ. + \item[100\% vs. 50\%] MZ twins share 100\% of their genes; DZ twins share + on 50\% + \end{description} + +\end{frame} + +\begin{frame}{ACE Assumptions II} + \begin{description} + \item[No \gxe] Genes do not interact with the environment. Example + failure: parents dress identical twins alike more often than fraternal + twin pairs, or co-enroll + them in activities more often. + \item[No \rge] Shared genes do not correlate with environmental stimuli. + Example failure: identical twins both join student council more often + than fraternal pairs. + \end{description} + These combine to form the \emph{Equal Environments Assumption (EEA)} \citep{Suhay:2009fy}. +\end{frame} +
 @@ -60,24 +60,12 @@ % \end{frame} -\begin{frame}{ACE Model} - Decompose observed monozygotic ($\rmz$) and dizygotic ($\rdz$) correlations - into: - \begin{description} - \item[Genetic] $A = 2 (\rmz - \rdz)$ - \item[Environment] $C = 2 \rdz - \rmz$ - \item[Error] $E = 1 - \rmz$ - \end{description} - % CHECK THIS: By implication, $A^2 + C^2 + E^2 = 1$. - By design $A + C + E = 1$. Variations fit using SEM and Bayesian methods. -\end{frame} - -\begin{frame}{Heritability} +\begin{frame}{Behavior Genetics} \begin{itemize} - \item $A$ is labeled heritability'' - \item Interpreted as the percentage of observed variance accounted for be + \item Goal: estimate percentage of observed variance accounted for by genetics. - \item Some examples: ideology \citep{ALFORD:2005kx}, vote turnout + \item Classical Twin Design widely used + \item Some political examples: ideology \citep{ALFORD:2005kx}, vote turnout \citep{FOWLER:2008rt}; political attitudes \citep{Eaves:2008ve} \item Three Laws of Behavior Genetics \citep{Turkheimer:2000uq} \begin{enumerate} @@ -88,27 +76,6 @@ \end{itemize} \end{frame} -\begin{frame}{ACE Assumptions I} - \begin{description} - \item[Randomized] Twin pairs are randomly assigned to be MZ or DZ. - \item[100\% vs. 50\%] MZ twins share 100\% of their genes; DZ twins share - on 50\% - \end{description} - -\end{frame} -\begin{frame}{ACE Assumptions II} - \begin{description} - \item[No \gxe] Genes do not interact with the environment. Example - failure: parents dress identical twins alike more often than fraternal - twin pairs, or co-enroll - them in activities more often. - \item[No \rge] Shared genes do not correlate with environmental stimuli. - Example failure: identical twins both join student council more often - than fraternal pairs. - \end{description} - These combine to form the \emph{Equal Environments Assumption (EEA)} \citep{Suhay:2009fy}. -\end{frame} - \begin{frame}{Sidebar: Variables in a Randomized Trial} \begin{description} \item[Treatments] Randomly applied (i.e. $Z \perp Y_0,Y_1$); have causal interpretation