Non-parametric tests


New Member
Hello everyone,

I am confused with two things it would be great if anyone could help.

1: Does the character type in a video game influence the duration of gaming (in minutes)? Each participant played a video game four times with different characters. Those characters were warrior, healer, bard, and scout. and each participant’s gaming duration was measured in minutes under four different gaming characters.

About the Friedman test, DV should be ordinal. if interval or ratio, should be converted to ordinal. that is as far as what ı know. in that case, should ı assign numbers to how long people played that video game depending on different characters starting from the lowest score(12 mins=1, next 35 mins=2 and so on) or ı should just use the dataset in minutes as it is and conduct the friedman test?

2: Does the accessory of a dog influence the placement in a dog cuteness contest? The accessory had two levels: ribbon and collar. The cutest dog was given the first place, the second cutest dog was given the second place etc. Hence, a lower rank indicates more cuteness. The ranking of each dog and their accessory was listed in the table below. Each line is a different dog.

I conducted a wilcoxon rank-sum test. Since a less cute dog with collar got lets say rank 19 between 20 dogs. since we should assign numbers from the lowest to highest starting with 1, should ı reverse the rankings and give that dog 2 since it became the second least cute dog before conducting wilxocon rank-sum test?

Thank you!