Relying on estimates of variance explained to assess the practical significance of research findings is now common practice. It is suggested that such estimates can offer an inaccurate picture, often underestimating the practical significance of statistically small effects. As an example, research on employment discrimination indicates that in non-traditional work settings women are generally judged less favorable than men. However, because the amount of variance due to sex is typically quite modest, the actual importance of sex bias effects has been questioned. Two numerical examples with 600 men and 600 women, and with 100 men and 100 women, respectively, demonstrate that even very small amounts of sex bias in hiring decisions and performance evaluations can have profoundly negative consequences for women. Computer simulations confirm this conclusion. It is hoped that researchers will not automatically discount the practical consequences of statistically small effects. Five tables and one figure illustrate the analyses. (Contains 32 references.) (SLD)