Dear all,
I would like to match (e.g. Mahalanobis match) observations that are nested in two cross-classified groups, Group1 (1-40) and Group2 (A-N).
In particular, observations are clustered in Group1 and these are are located in all Group2 clusters. Group2 clusters are the reference groups. I would like to match all Group1 observations to the observationally equivalent/close observations in Group2 based on a set of covariates.
I am then after assessing whether an aggregate level Group1 independent variable influences a binary outcome when observations nested in Group1-40 are compared only to the respective observations of reference GroupA-N that are Mahalanobis matched (and accounting for aggregate GroupA-N effects).
Would you be able to help me with this undertaking? Is this possible? I have had a look at e.g. psmatch2 (scc) but I am not sure I am able to identify any solution. The typical case only helps with the ATE for Group1 vs. Group2 if I am not mistaken?
Many thanks!
I would like to match (e.g. Mahalanobis match) observations that are nested in two cross-classified groups, Group1 (1-40) and Group2 (A-N).
In particular, observations are clustered in Group1 and these are are located in all Group2 clusters. Group2 clusters are the reference groups. I would like to match all Group1 observations to the observationally equivalent/close observations in Group2 based on a set of covariates.
I am then after assessing whether an aggregate level Group1 independent variable influences a binary outcome when observations nested in Group1-40 are compared only to the respective observations of reference GroupA-N that are Mahalanobis matched (and accounting for aggregate GroupA-N effects).
Would you be able to help me with this undertaking? Is this possible? I have had a look at e.g. psmatch2 (scc) but I am not sure I am able to identify any solution. The typical case only helps with the ATE for Group1 vs. Group2 if I am not mistaken?
Many thanks!
