Of the three predictivemodels devel

Of the three predictivemodels developed for short-term employee commitment, this model showed that respondents between the ages of 31 and 50 had the highest probability of remaining with the employer for at least six months.PðY ¼ 1Þ¼e½2.5135þð−0.0860ÞðRwCÞþ0.1727ðEBÞ
1þe½2.5135þð−0.0860ÞðRwCÞþ0.1727ðEBÞ
ð4Þ The predictive model equation for age 3 (ages 51 and older) used the regression coefficient for age 3 versus age 1. In this predictive model, the regression coefficient for age 3 was actually a negative number (−0.5836). Thus, respondents who were 51 years or older were less likely to stay with their current employer for six months compared to those aged 50 and younger. This was reflected in the intercept for Eq. (5), which was 1.3423. It was derived from theinterceptinEq.(3)(thepredictivemodelforage1),andwasthe difference between the age 1 coefficient of 1.9259 and the age 3 coefficient of−0.5836. The rest of the equation remained the same as the first predictive model [Eq. (3)].PðY ¼ 1Þ¼e½1.3423þð−0.0860ÞðRwCÞþ0.1727ðEBÞ
1þe½1.3423þð−0.0860ÞðRwCÞþ0.1727ðEBÞ
ð5Þ Table 8lists the odds ratios associated with the predictivemodel short-term employee commitment. Age 1 (age 30 and less) is at the intercept of the predictive model equation. Age 2 respondents’ (ages 31 through 50) probability of responding “likely” to the dependent variable short-term employee commitment increased by 1.8 times, and age 3 (age 51 and older) respondents’ probability of a “likely” response increased by 0.56 times more than the response values from the age 1 group. Respondents in age 2 were morethan3timesmorelikelytorespond“likely”tothequestionof short-term commitment than those in age 3. The predictive model was applied to the validation data set and predicted the validation data set correctly 82.6% of the time.