A dataset simulated to exemplify the use of the multilevelIV()
function.
It has 2767 observations, clustered into 40 level-three variables and 1347 observations at level two. The endogenous regressor is X15
with a true
coefficient value of -1.
data("dataMultilevelIV")
A data frame with 2767 observations clustered into 40 level-three variables and 1347 level-two variables.
y
a numeric vector representing the dependent variable.
X11
a level-one numeric vector representing a categorical exogenous variable with true parameter value equal to 3.
X12
a level-one numeric vector representing a binomial distributed exogenous variable with true parameter value equal to 9.
X13
a level-one numeric vector representing a binomial distributed exogenous variable with true parameter value equal to -2.
X14
a level-two numeric vector representing a normally distributed exogenous variable with true parameter value equal to 2.
X15
a level-two numeric vector representing a normally distributed endogenous variable, correlated with the level-two errors. It true parameter value equals to \(-1\) and it has a correlation with the level two errors equal to 0.7.
X21
a level-two numeric vector representing a binomial distributed exogenous variable with true parameter value equal to -1.5.
X22
a level-two numeric vector representing a binomial distributed exogenous variable with true parameter value equal to -4.
X23
a level-two numeric vector representing a binomial distributed exogenous variable with true parameter value equal to -3.
X24
a level-teo numeric vector representing a normally distributed exogenous variable with true parameter value equal to 6.
X31
a level-three numeric vector representing a normally distributed exogenous variable with true parameter value equal to 0.5.
X32
a level-three numeric vector representing a truncated normally distributed exogenous variable with true parameter value equal to 0.1.
X33
a level-three numeric vector representing a truncated normally distributed exogenous variable with true parameter value equal to -0.5.
SID
a numeric vector identifying each level-three observations.
CID
a numeric vector identifying each level-two observations.
Raluca Gui raluca.gui@business.uzh.ch