chi square

Question

In a Chi-square question you are usually given observed and you need to fill in the rest: expected, expected^2, etc. (Or you need to find the observed. But in this example observed is given) To get expected you multiply by a ratio. At times this problem is a recombinant (R) vs non-recombinant (NR) problem where you add the R with R and the NR with the NR. Why? And when do you know to do this? What if you have a 3:1 problem? Do you take all three groups out of four and group as one vs the small one group?

Answer 1

This sounds like it is testing whether a trait is dominant, recessive or neither. In the case of neither, you would only find that NR traits show up with NR crossed with NR while recombinant traits only show up when recombinants are crossed. (when NR and R cross random stuff happens.)

 

If we expect one group to be dominant then anything with R will show the trait (RxNR,NRxR,RxR) while the NRxNR case will not, thus the 3:1 ratio. It just depends on your null hypothesis.