>>34Britain is about 95% white. Globally, about 92% of programmers are male. Assuming these percentages are representative of the total British programming population, then one would expect that a representative sample of British programmers would have a (1-(0.95^N)) chance of being 100% white, and a (1-(0.92^N)) chance of being 100% male, where N is the size of the sample.
The article indicates a sample size of N = 15. This sample is reported to consist entirely of white males. The probability of this sample being all-male is 54%. The chance of this sample being all-white is 71%. The chance of the sample being BOTH all-white and all-male is 38%. Which means, there is a 62% chance of discrimination being involved somewhere along the line.
0.95^N is the probability of the N-sized group all being white. 0.92^N is the probability of being all male. (0.95*0.92)^N is probability of all white males. (1 - (0.92*0.95)^N) is the probability of at least one non-white or non-male in the group, which is 86.7% where N = 15.