Hello Statalist,

this is my first post and I hope you can help me interpreting my regression coefficients. I have estimated the following two OLS models (using Stata command reg):

1. CAR = a + ß*Satisfaction + ß*Controls + e

CAR stands for the cumulative abnormal return and ranges between -0.09 and 0.10. Satisfaction is a percentage and is calculated as votes for divided by the votes cast, with a standard deviation of 12.31%. The ß-coefficient on Satisfaction is -0.346. Therefore my interpretation is as follows:

A one percentage point increase in Satisfaction is associated with an estimated decrease in CAR of 0.346 percentage points.
Or alternatively, a one standard deviation (12.31%) increase in Satisfaction is associated with an estimated decrease in CAR of 4.26 percentage points (=-0.346*0.1231)


2. |CAR| = a + ß*ln(Number of News) + ß*Controls + e
In this model, I use the absolute value of CAR as the dependent variable and ln(Number of News) is the independent variable of interest. The coefficient on ln(Number of News) is -0.035. So my interpretation is as follows:

A 1% increase in the number of news is associated with an estimated decrease of the unsigned CAR of 0.0348 percentage points (= -0.035*ln(1.01) = 0.000348)
==> But this is just an approximation as it is a level-log regression.

I hope that my interpretation is correct. Thank you in advance for your comments.

Best regards

Dario