2. The researcher also believes that ethnicity might affect the probability of arrest and obtains the following data non blaclq"non mixed racefnon Black Hispanic Hispanic Hispanic number arrested T54 534 25 1252 number not arrested 1384- 1210 49 2399 Ic) How would we test whether the probability of being arrested is affected independently by both gender and ethnicity? What about testing for whether the effect of gender is different for different ethnic groups? [10 marks] (I) Why might this analysis give us the wrong answer about the relationship between ethnicity, gender and the probability of arrest":I [5 marks]