在本文中,利用非中心卡方分配的方法及Patnaik(1949)、Zar(1978)、 Wilson-Hilferty(1931)所提出的近似式,以區間估計的觀念探討製程能力指標,並考慮在α風險及檢定力的條件下,分別推導出C/sub pm/指標的最小臨界值和C/sub pp/指標的最大臨界值,以及相對的最適樣本數,以做為製程能力檢定的決策準則,且提供了最適樣本數的選取參考。 In this paper, we use the approach of the non-central chi-square distribution and some approximated formulae, proposed by Patniak(1949), Zar(1978) and Wilson-Hilferty(1931), to derive the minimum critical values of the process index C/sub pm/ and the maximum critical values of the process index C/sub pp/ under some given values of the α-risk and the power. The corresponding optimal sample sizes are also determined.