Paired t-Test by comparing 2 set of mean/median

[megaproto]

Hello lets say im doing a test measuring people heartbeat before and after various activities. I have 10 participants and 10 types of activities. The activities are basically playing video games. So first all 10 participants play 10 types of games (eg sports, war etc) and the games are muted which mean participants will not hear any sound from the game. I then record the participants heartbeat for each game. For the next set the participants plays the same games but with the sound turn on and the heartbeat is recorded.

So i come up with a hypothesis that the heartbeat of participants should increase when playing game with sounds in comparison to games without sounds

Hnull : heartbeat = 0 (the heartbeat doesnt increase)
H1: heartbeat = not 0 (heartbeat increase)

is it ok, rather than measuring the heartbeat of participants for each type of game (before and after), instead, i take the mean (normally distributed) or median (due to outlier) of the heartbeat for each game for before and after then compare the two instead?

so
_________________________________|G1.............G10
heartbeat mean/median before sound | X1..............X10
heartbeat mean/median after sound | Y1..............Y10


thanks in advance

 

[royhaas]

You would be better off if everyone played the same game.