Nitpick here. > Referrals are the minority of applicants but are five times more likely to be hired. From the article: P(referral) = 0.06 P(non-referral)

Nitpick here. > Referrals are the minority of applicants but are five times more likely to be hired. From the article: P(referral) = 0.06 P(non-referral) = 0.94 P(referral|hired) = 0.29 P(non-referral|hired) = 0.71 What we're looking for is P(hired|referral) / P(hired|non-referral) = ? So reversing the two known conditionals, and dividing to eliminate the unknown probability for P(hired): P(hired|referral) = P(referral|hired) * P(hired) / P(referral) P(hired|non-referral) = P(non-referral|hired) * P(hired) / P(non-referral) P(hired|referral) / P(hired|non-referral) = P(referral|hired) / P(non-referral|hired) * P(non-referral) / P(referral) = 0.40845 * 15.667 = 6.3991 Referrals are actually 6.4 times as likely to be hired. This is why stat and math classes are important, journalists.