In order to observe attitudinal differences of employees, if there is any, within HPWL and HPSL group we classified our respondents of each group into managerial and technical categories and the distributions of their prime reasons behind leaving company are presented in table-2 and table-3.
It is evident (see table-2) that the both the managerial and technical category of employees of HPWL group revealed the same reasons behind their leaving the last preceding company.
Two logistic regression models have been run separately, one for the HPWL group and the other for the HPSL group of employees.
It is evident from the classification table (see table M8) that by adding the explanatory variables we can now predict HPWL and HPSL employees' propensity to change job with 95.4 percent and 94.3 percent respectively accuracy.
In our case the model chi square for HPWL group of employees has 6 degrees of freedom, a value of 252.124 and a probability of p < 0.000 and HPSL group has 6 degrees of freedom, a value of 186.095 and a probability of p < 0.000.
Here it is indicating 68.7 percent (in case of HPWL group) and 69.0 percent (in case of HPSL group) of the variation in the dependent variable is explained by the logistic models.
Here in our models H-L statistic have significance of .959 for HPWL group and .995 for HPSL group of employees which means that these are not statistically significant and therefore our models are quite good fit (see table M7).