I have collected data that represent stimulation levels in three different time points ($T_0$, $T_1$ and $T_2$) using three different stimulants ($S_1$, $S_2$ and $S_3$) and a control ($C$). I have tested all three stimulants with three different subjects ($J_1$, $J_2$ and $J_3$) and for each subject and stimulant/ control I have three numbers that represent stimulation levels in $T_0$, $T_1$ and $T_2$ time points.
How can I test the hypothesis that a certain stimulant shows significant higher level than the control? How can I test the hypothesis that all three stimulants shows a significant/ insignificant difference?
The data is not normally distributed so I would have to use non-parametric tests. I am lost to whether it should be paired/unpaired or to use a regression or something else. Can anybody help?