Hi all,
I ran a probit model. My dependent variable is an acquisition dummy that takes the value one if the firm make an acquisition and zero otherwise. I ran the model twice. In the first model my independent variable is conservatism index (continuous variable) measured as [(number of negative words/total words) *100]. The marginal effect for this independent variable is - 0.022. I interpret as follows
A one-unit increase in conservatism index reduces the likelihood of an acquisition by 2.2 percentage points. (Actually, I am little bit confused in calculating the unit of conservatism index, I do appreciate your help in this is it a one-percentage point of one percent or it is better to keep as one unit).
Do you think it is better to say “The marginal effects means that a one-percentage point increase in conservatism index reduces the likelihood of an acquisition by 2.2 percentage points”
In the second model my independent variable is a dummy variable in particular highly conservative firms that takes the value one if the conservatism index for the firm is higher than the sample median and zero otherwise. The marginal effects for this variable is -0.013. I interpret as follows.
Moving from a less conservative position to a highly conservative one (the move from 0 to 1) reduces the likelihood of an acquisition by 1.3 percentage points.
My questions;
First, are these interpretations correct?
Second, how I could define the unit increase for the conservatism index in the first model?
I do really appreciate your help.
I ran a probit model. My dependent variable is an acquisition dummy that takes the value one if the firm make an acquisition and zero otherwise. I ran the model twice. In the first model my independent variable is conservatism index (continuous variable) measured as [(number of negative words/total words) *100]. The marginal effect for this independent variable is - 0.022. I interpret as follows
A one-unit increase in conservatism index reduces the likelihood of an acquisition by 2.2 percentage points. (Actually, I am little bit confused in calculating the unit of conservatism index, I do appreciate your help in this is it a one-percentage point of one percent or it is better to keep as one unit).
Do you think it is better to say “The marginal effects means that a one-percentage point increase in conservatism index reduces the likelihood of an acquisition by 2.2 percentage points”
In the second model my independent variable is a dummy variable in particular highly conservative firms that takes the value one if the conservatism index for the firm is higher than the sample median and zero otherwise. The marginal effects for this variable is -0.013. I interpret as follows.
Moving from a less conservative position to a highly conservative one (the move from 0 to 1) reduces the likelihood of an acquisition by 1.3 percentage points.
My questions;
First, are these interpretations correct?
Second, how I could define the unit increase for the conservatism index in the first model?
I do really appreciate your help.