To test hypotheses about the likelihood of LTIP use
following adoption, for those firms adopting LTIPs, we used
logistic regression analysis. Ordinary least squares (OLS)
analysis is inappropriate when the dependent variablei s
categorical, because OLS assumes a linear additive model
with normally distributed error terms, while the true
probability model is nonlinear with binomially distributed
errors (Hosmer and Lemeshow, 1989). Finally, we used OLS
regression analysis to test predictions about the magnitude
of LTIP grants, for those firms making such grants.
To test hypotheses about the likelihood of LTIP usefollowing adoption, for those firms adopting LTIPs, we usedlogistic regression analysis. Ordinary least squares (OLS)analysis is inappropriate when the dependent variablei scategorical, because OLS assumes a linear additive modelwith normally distributed error terms, while the trueprobability model is nonlinear with binomially distributederrors (Hosmer and Lemeshow, 1989). Finally, we used OLSregression analysis to test predictions about the magnitudeof LTIP grants, for those firms making such grants.
การแปล กรุณารอสักครู่..
