Design

You need to seperate your independent variable into levels. Eg a low stressful, medium stressful, high stressful and no stress conditions.

Thanks a lot for your reply.

In my experiment I used one leve, which you can say high stressful. Is that fine if I use one level?

Of course not. That's not an experiment.

What would you compare that result to in order to show that stress caused the headache? You need a control.

I used somthing like a scale to rate the headache at the beginning of the task and at the end. Is that acceptable to be a control?

It is a 0 (no headache)-5 (severe headache) scale, should I change it to 0-4 scale?

You still really should be using a control group that goes through time 1 and time 2 measurements. Give them a non stressful task to do.

Changing the arbitrary numbers of your scale makes no difference to the issue that I posted (either in this thread or one of your many others).