Hello,
I am attempting to determine if there is a relationship between job satisfaction and working hours. To do so, I have been employing a hierarchical regression to determine if there is a non-linear relationship between the two variables using the following code:
reg job_satisfaction age monthlypay education
reg job_satisfaction HW_centered age monthlypay education
reg job_satisfaction HW_centered HW_centeredsq age monthlypay education
However, when I try to include employment type (wage employed vs. self employed) as a moderating variable, I encounter collinearity between HW_centered and my binary self_emp moderating variable. I have tried transforming HW_centered by log-ing the values and introducing polynomials, but I keep encountering the same problem. I have recoded my variables, but since self_emp is a binary variable I am a little bit lost on what else I can possibly try to resolve collinearity.
This is the output when I include self_emp as a variable:
Any advice would be greatly appreciated
Regards,
Ayyan Tareen
I am attempting to determine if there is a relationship between job satisfaction and working hours. To do so, I have been employing a hierarchical regression to determine if there is a non-linear relationship between the two variables using the following code:
reg job_satisfaction age monthlypay education
reg job_satisfaction HW_centered age monthlypay education
reg job_satisfaction HW_centered HW_centeredsq age monthlypay education
However, when I try to include employment type (wage employed vs. self employed) as a moderating variable, I encounter collinearity between HW_centered and my binary self_emp moderating variable. I have tried transforming HW_centered by log-ing the values and introducing polynomials, but I keep encountering the same problem. I have recoded my variables, but since self_emp is a binary variable I am a little bit lost on what else I can possibly try to resolve collinearity.
This is the output when I include self_emp as a variable:
| . reg job_satisfaction HW_centered HW_centeredsq | self_emp age monthlypay education |
| note: self_emp omitted because of collinearity. | |
| Source SS df MS | Number of obs = 13,241 |
| F(5, 13235) = 6.53 | |
| Model 14.2207733 5 2.84415467 | Prob > F = 0.0000 |
| Residual 5764.34323 13,235 .435537834 | R-squared = 0.0025 |
| Adj R-squared = 0.0021 | |
| Total 5778.56401 13,240 .436447432 | Root MSE = .65995 |
| job_satisfa~n Coefficient Std. err. t | P>t [95% conf. interval] |
| HW_centered -.0016537 .0006207 -2.66 | 0.008 -.0028703 -.000437 |
| HW_centeredsq .0000213 .0000274 0.78 | 0.436 -.0000323 .0000749 |
| self_emp 0 (omitted) | |
| age .0004921 .0004349 1.13 | 0.258 -.0003604 .0013446 |
| monthlypay .0000226 4.39e-06 5.16 | 0.000 .000014 .0000312 |
| education .0156488 .0084475 1.85 | 0.064 -.0009094 .032207 |
| _cons 2.593788 .0243984 106.31 | 0.000 2.545964 2.641613 |
| . | |
| end of do-file | |
Regards,
Ayyan Tareen
Comment