對於可重覆讀寫光碟（Compact Disc-Rewrittable：CD-RW）之重覆讀寫壽命（又稱重覆讀寫度；Cycliability）測試，做全數測試不但費時且浪費成本，此種測試為破壞性測試且需花約2.5天/片。本文之研究重點在於推導出CD-RW重覆讀寫度測試，在生產線上每次抽樣測試所需之間隔批量大小（l）及每次抽樣樣本數（n）。逐次抽樣法被使用於本文之目的在於減少n值。而逐次抽樣法之往下單向規格的決定是根據逐次抽樣機率比、α與β（即型I和型II誤差之風險機率）、可重覆讀寫度（或燒錄）次數（t）、可重覆讀寫次數之公差（T）及測試重覆讀寫度母全體之標準差（σ）等。此單向規格可以用來決定CD-RW之重覆讀寫度所需之n值。理論上，逐次抽樣法之平均樣本數（ANSS）亦可由上述之α,β,t,T及σ而求得。另外，本文依據卡方分配之自由度、（1-σ）信賴水準及樣本變異S2來決定出其估計母體變異數σ2之範圍（即Estimated Range ofσ2；ERσ2）。並利用卡方近似值反函數式在不同信賴水準下，v = (m – 1)，ERσ2與n之關係來推導出在生產線上抽樣時所需之抽樣間隔批量大小值。

　　　　　100% of Compact Disc-Rewrittable (CD-RW) life testing (or cycliability testing) is time-consuming and costly. Since this kind of testing is a destructive test and need spend 2.5 ~ 3 days/piece. Therefore, the research focus is to derive the intermittent batch sizes (l) and sample sizes (n) in the production line for every cycliability testing of CD-RW. It is noted that sequential test method reduces the sample sizes, n, greatly. A downward specification of sequential tests based on the sequential probability ratio test, risk levels of α andβ (ie., Type I and Type II error), estimated written-counts, t, of CD – RW, and the tolerance, T, of written-count, and the population standard deviation, α, of written-count have been applied to determine the sample sizes, n, for cycliability test of CD-RW. Theoretically, the average number of sample sizes (ANSS) for sequential test required determining the cycliability of CD-RW has also been derived based onα, β, t, T andσ. Furthermore, the degree of freedom in chi-square distribution, (1 –σ) level of confidence and sampling variance S2 have been used to determine the estimated range ofσ2 (ERσ2). An inverse chi-square approximation have also been used to determine the relationship between v = (m – 1), ERσ2 and n at different levels of confidence to determine the batch sizes, l, in a production line.