>Transforming weight into a binary variable with an arbitrary separation point is not sound statistical practice But the separation is not arbitrary. "Light en

>Transforming weight into a binary variable with an arbitrary separation point is not sound statistical practice But the separation is not arbitrary. "Light enough to wield by women" is the important statistic here, so the value in question is binary by nature. Weight in isolation simply isn't what is being considered here.