Prediction models and correlations
Forgive if I am not a seasoned statistician, I am starting out since some months. It's very difficult to starting out, I think you know it.
I don't know many terms, and what it's super-difficult is also to learn JASP language and tools in JASP, how do they work.
I am mining data on tennis. I have a lot of data about a tennis player.
I want to find some correlations.
I target tennis players. I want to know if there is a correlation between a tennis player's victory and the number of ACEs made during his last 30 matches. Can you help me write this in JASP?
What distribution should I choose? Poison? What to put on the axis of the abscissa and what to put on the axis of the ordinates? What are the variables in this case? How to write this in statistical terms? what distribution to use?
ALSO: How to have a model that predicts what would happen if the number of aces were below a certain threshold? Thank you so much if you could answer to me
Thank you
Comments
Hi,
I moved this discussion to the correct subforum.
Good luck,
Eduard
no answer?
Hi Luchins,
I was a little confused by Eduard's message above. In general, if you want to compute a correlation between victory and the number of aces you have to take into account that victory is a binary variable. It seems you would need the point-biserial correlation coefficient for this purpose. We don't have this in JASP (yet) but I'm sure there are R packages that compute this for you, together with documentation. The average number of aces in 30 matches you can safely treat as a continuous variable I would say.
Cheers,
E.J.
Hi EJ,
Is point-biserial correlation coefficient still not avaiable in JASP?
Thank you,
Sameha
I am surprised to see that it is not. If you add it as a feature request on our GitHub page I'll give it some priority.
Noted! Thank you EJ