Your Power Ranking is calculated using the Wilson score Lower bound formula, a binomial proportion confidence interval which is an interval estimate of a success probability {p} when only the number of experiments {n} and the number of successes {n_s} are known. In this case, {p} is your power ranking, {n} is the number of questions answered and {n_s} is the % correct on those questions. The formula is

Power Ranking = (p + z²/2n − z·√(p(1−p)/n + z²/4n²)) / (1 + z²/n)

Huh? In short, The Wilson lower bound is a way of scoring things that automatically punishes small sample sizes, so a handful of perfect scores can't unfairly beat a large pile of very-good scores. It provides an objective way to rank users even when the number of questions answered is not the same.