Thesis Help Request - Hypothesis Testing

#1
Hi I am doing an event study and my null hypothesis is CAAR=0 and alternative is CAAR does not equal 0 (2 tailed t-test). CAAR is cumulative average abnormal returns. Can a two tailed test be used to give directional conclusions, e.g. if the CAAR value is 1.5% and is statistically significant at say 5% significance level for a two tailed test, can I talk about the value being positive, so the affect of X on firm stock prices has a significant increase of 1.5%
 
#2
You would interpret the result with respect to the alternative hypothesis. "There is evidence that the average CAAR is not equal to 0." If you want to speak about direction, then the alternative hypothesis should be one-sided. What statistical model are you using? Are the assumptions of the model met? How is the response variable measured?
 
#3
You would interpret the result with respect to the alternative hypothesis. "There is evidence that the average CAAR is not equal to 0." If you want to speak about direction, then the alternative hypothesis should be one-sided. What statistical model are you using? Are the assumptions of the model met? How is the response variable measured?
A typical market model is used. Is it still possible to comment on the sign of the CAAR value or would it be incorrect practice to do so? A large number of previous papers use a two sided test but comment on directional values CAAR. E.g. CAAR significantly different from 0 and 1.5% gain seen following the event...
 
#4
I'm not familiar with the subject. I suppose if it's in reputable literature that it would be fine. Perhaps others on this site will have more to contribute.