Continue on the life of asymmetry. Following is the summary statistics of the admission data at Cambridge in 1996 in science, technology, engineering and mathematics (STEM). What is your conclusion?
| Accepted (women) | Accepted (men) | |
| Total | 274 | 584 |
It doesn’t look good, isn’t it? A clear case of gender discrimination. Now, look at the data further.
| Applied (women) | Accepted (women) | % | Applied (men) | Accepted (men) | % | |
| Total | 1184 | 274 | 23.1 | 2470 | 584 | 23.6 |
No real difference in the percentage accepted.
Go further deep
| Applied (women) | Accepted (women) | % | Applied (men) | Accepted (men) | % | |
| Computer Science | 26 | 7 | 27 | 228 | 58 | 25 |
| Economics | 240 | 63 | 26 | 512 | 112 | 22 |
| Engineering | 164 | 52 | 32 | 972 | 252 | 26 |
| Medicine | 416 | 99 | 24 | 578 | 140 | 24 |
| Veterinary Medi cine | 338 | 53 | 16 | 180 | 22 | 12 |
| Total | 1184 | 274 | 23 | 2470 | 584 | 24 |
Now, this is interesting! Every department accepted a higher proportion of women who applied, yet the overall percentage favoured men. Known as Simpson’s paradox, this reversal of interpretation after accounting for confounding factors is something we should always pay attention to. In this case, women preferred more competitive departments with lower acceptance rates, whereas more men opted for engineering, which had better acceptance rates.
