# fitting suitable regression model to identify predictors from contingency table

so a study was conducted for some game and the probability of success of the game was noted in a contingency table.

I have a 5x5 contingency table which is age group by task difficulty(split into easy and hard). difficulty is split into experiment 1 and experiment 2.

so the 5 columns are:age group, experiment1 easy,experiment 1 difficult, experiment 2 easy, experiment 2 difficult and the values are all probabilities.

I do apologise as I am unsure how to encode this on latex.

Now I think I have to fit a generalised linear model to see what covariates are significant and then perform model selection but I have never approached it from a contingency table and im unsure how to go about tackling this. should I copy the Values $$X_{1,1}$$ to $$X_{5,5}$$ into R and create my own data set and work off of that?