Hi, I'm just wondering if anyone here can help me.
I'm making a model for my logistic regression. For the variables, I'm using 1 Dichotomous Dependent Variable (Financial Distresss, Y) and 2 Continuous Independent Variable (Corporate Governance Score, X and VAIC Score, Z). I'm using logistic regression in order to find the relation and its significance of X to Y, Z to Y and interaction between X and Z to Y. The sample amount I'm using is n=136 Mathematically speaking, it is expressed as:
log(Y/1Y)=a + bX + cZ + dX*Z
However, when I ran the logistic regression in SPSS 25, my HosmerLemeshow GoodnessofFit Significance value is 0.00 , which is interpreted as a lackoffit in the model.
Is there anything that I can do to my data in order to improve the significance value of my HosmerLemeshow (HL) in SPSS?
I've searched in the web and there aren't many detailed solutions about it, some even dismissed the idea of using HL test as a means to see whether the model is fit or not. But because of my sample size, which is relatively small, I really need to carryout this test.
I've attached the data and its output in this thread that can be opened by SPSS if you're interested in seeing it. Thank You!
I'm making a model for my logistic regression. For the variables, I'm using 1 Dichotomous Dependent Variable (Financial Distresss, Y) and 2 Continuous Independent Variable (Corporate Governance Score, X and VAIC Score, Z). I'm using logistic regression in order to find the relation and its significance of X to Y, Z to Y and interaction between X and Z to Y. The sample amount I'm using is n=136 Mathematically speaking, it is expressed as:
log(Y/1Y)=a + bX + cZ + dX*Z
However, when I ran the logistic regression in SPSS 25, my HosmerLemeshow GoodnessofFit Significance value is 0.00 , which is interpreted as a lackoffit in the model.
Is there anything that I can do to my data in order to improve the significance value of my HosmerLemeshow (HL) in SPSS?
I've searched in the web and there aren't many detailed solutions about it, some even dismissed the idea of using HL test as a means to see whether the model is fit or not. But because of my sample size, which is relatively small, I really need to carryout this test.
I've attached the data and its output in this thread that can be opened by SPSS if you're interested in seeing it. Thank You!
Attachments

17.3 KB Views: 0