The data just represents the population it was taken from. The problem is the inference the author tries to imply.