4 ms·

100 is such a small N that 50-50 could just be rounding noise, I'd think. That's why I'm glad that they care about such things, but it doesn't suggest to me tha

100 is such a small N that 50-50 could just be rounding noise, I'd think. That's why I'm glad that they care about such things, but it doesn't suggest to me that they are truly unbiased in their selection process. The only thing that makes sense to draw conclusions about wrt diversity are the make up of a) large corporations b) university output and c) the distribution of groups throughout start-ups as a whole. They might have beat the average, but without more data, it really could be anything else other than what they posit they did.

SuperImprobable 10y ago
The 95% confidence interval would be about 40-60%. https://www.wolframalpha.com/input/?i=binomial+confidence+interval&rawformassumption=%22FSelect%22+-%3E+%7B%7B%22BinomialInterval%22%7D%7D&rawformassumption=%7B%22F%22,+%22BinomialInterval%22,+%22c%22%7D+-%3E%220.95%22&rawformassumption=%7B%22F%22,+%22BinomialInterval%22,+%22n%22%7D+-%3E%22100%22&rawformassumption=%7B%22F%22,+%22BinomialInterval%22,+%22phat%22%7D+-%3E%220.5%22 https://www.wolframalpha.com/input/?i=binomial+confidence+in...