1

Assume we have a lottery machine where you press a button and it returns one of 5 motifs in one of 5 colours. Each of these also has a chance to be gilded.

Assume I have a dataset containing outcomes of this lottery machine. i.e. a list of motifs, their colours and whether or not it is gilded.

How would I go about estimating the probabilities of getting each combination? How would I determine whether gilding probability is independent of motif and/or color?

I'm very familiar with linear normal models, mme models and hypothesis testing in these models but I have no idea how to proceed with a problem like this.

Lundborg
  • 135
  • 6

1 Answers1

1

There is a well known test to check independence of two variables: Pearson chi-squared test of independence. You can use it to test if gilding and color are independent, and you can use it to test if gilding and motif are independent. I would also try to test if the combination of motif and color is independent from gilding.

Pere
  • 6,583