There is more detail required to suggest something. Is it the same sample? If so the independent sample t test would be viable to see if mean scores differ in a statistically significant way...
Maher, if those two samples seem to be representative to the student population you can run a t test (compare mean scores for stress) and then you can suggest if levels of stress differ significantly at time point 1 and 2. You can also run t test to see if there statistically significant differences across gender and other demographics.