There's a lot of "depends" here and I don't have enough information to help you draw a conclusion.
First things first:
Are you trying to see if some set of data is equal to, or different than, another set?
If equal, then you need to look at the 2-tail test
If different, then you need to look at the 1-tail test.
The T-value 8.9 is not sufficient to answer the question whether the sets of data are equal or different by itself; although yes, 8.9 is bigger than both 1.65 and 1.97.
You need to look at the P-Value to determine whether the difference is statistically significant.
Since you're using an Alpha coefficient of .05 (or said another way you want your answer with a 95% level of confidence), then the P-value must be less than .05 to be statistically significant.
If the P-Value for the tail type your testing is less than .05, then yes you should reject the null.