Hi Statalist members,
I'm running the ordered logit model and find struggling how to interpret the results of an interaction term of two continuous variables.
My dependent variable is debt maturity, which takes a value of 1 if a firm use long term debt, 2 if a firm only uses short term debt and 3 if a firm does not use any long term or short term debt in a given year. Although I know this order is quite ambiguous to Stata experts here ( and most of you will recommend me to use the multinomial logit, which seems not work with my dataset), please ignore that issue in this post.
My main variable of interest is firm cash flow volatility (CFV)
My model is Debt maturity = alpha + beta_1 * CFV + beta_2 * other firm-level control variables + beta_3 * country-level variables + error term
Now I would like to test if a country's banking system strengthen or weaken the influence of cash flow volatility on debt maturity.
The additional mode is alpha + beta_1 * CFV + beta_2 * other firm-level control variables + beta_3 * country-level variables + beta_4* bank*CFV) + error term
I searched and found a few useful posts in this forum, suggesting I should use the margin command to figure out the effect of an interaction term and I can't interpret the coefficient of CFV as in the model without interactions any more. However, I still don't know how to interpret the results after getting the marginal effect of CFV and interaction.
Here are the cash flow volatility marginal effects ( on three categories of debt maturity)
Delta-method
dy/dx Std. Err. z P>z [95% Conf. Interval]
KS_oi_5_w01
_predict
1 -.0011547 .0002561 -4.51 0.000 -.0016566 -.0006528
2 .0003506 .0000782 4.48 0.000 .0001972 .0005039
3 .0008041 .0001782 4.51 0.000 .0004548 .0011535
And here are the interaction of the banking system and cash flow volatility marginal effects
Delta-method
dy/dx Std. Err. z P>z [95% Conf. Interval]
baKS
_predict
1 -5.32e-06 2.52e-06 -2.11 0.035 -.0000103 -3.70e-07
2 1.61e-06 7.67e-07 2.10 0.035 1.10e-07 3.12e-06
3 3.70e-06 1.76e-06 2.11 0.035 2.58e-07 7.15e-06
Any help would appreciate.
Regards.
Huyen
Huyen
I'm running the ordered logit model and find struggling how to interpret the results of an interaction term of two continuous variables.
My dependent variable is debt maturity, which takes a value of 1 if a firm use long term debt, 2 if a firm only uses short term debt and 3 if a firm does not use any long term or short term debt in a given year. Although I know this order is quite ambiguous to Stata experts here ( and most of you will recommend me to use the multinomial logit, which seems not work with my dataset), please ignore that issue in this post.
My main variable of interest is firm cash flow volatility (CFV)
My model is Debt maturity = alpha + beta_1 * CFV + beta_2 * other firm-level control variables + beta_3 * country-level variables + error term
Now I would like to test if a country's banking system strengthen or weaken the influence of cash flow volatility on debt maturity.
The additional mode is alpha + beta_1 * CFV + beta_2 * other firm-level control variables + beta_3 * country-level variables + beta_4* bank*CFV) + error term
I searched and found a few useful posts in this forum, suggesting I should use the margin command to figure out the effect of an interaction term and I can't interpret the coefficient of CFV as in the model without interactions any more. However, I still don't know how to interpret the results after getting the marginal effect of CFV and interaction.
Here are the cash flow volatility marginal effects ( on three categories of debt maturity)
Delta-method
dy/dx Std. Err. z P>z [95% Conf. Interval]
KS_oi_5_w01
_predict
1 -.0011547 .0002561 -4.51 0.000 -.0016566 -.0006528
2 .0003506 .0000782 4.48 0.000 .0001972 .0005039
3 .0008041 .0001782 4.51 0.000 .0004548 .0011535
And here are the interaction of the banking system and cash flow volatility marginal effects
Delta-method
dy/dx Std. Err. z P>z [95% Conf. Interval]
baKS
_predict
1 -5.32e-06 2.52e-06 -2.11 0.035 -.0000103 -3.70e-07
2 1.61e-06 7.67e-07 2.10 0.035 1.10e-07 3.12e-06
3 3.70e-06 1.76e-06 2.11 0.035 2.58e-07 7.15e-06
Any help would appreciate.
Regards.
Huyen
Huyen
Comment