tncI was very close to picking that as well. However, I ended not going with it because I can just control how many in each groups without picking at random. I would only think about picking at random if I don't want any bias in putting people on one side or another based off certain characteristics. Hence, picking at random would probably by chance put equal similar characteristics on each sides, rather than how many on each side. I hope that makes sense.+4
cbrelandI agree, I feel like randomization is better for balancing characteristics and potential confounders because it would be hard to take in any and all differences in people when hand picking the groups.+2
submitted by โanechakfspb(77)
I personally picked two groups with similar sample size, can someone explain how they answered this question? Thank you!