Schoonahd J W, Gould J D, Miller L A
a IBM Thomas J. Watson Research Center , Yorktown Heights , Now York , 10508 , U.S.A.
Ergonomics. 1973 Jul;16(4):365-379. doi: 10.1080/00140137308924528.
This paper describes the results of four experiments in a series aimed at understanding and improving visual inspection in general and of small integrated circuits (i.e. 'chips') in particular. Stimuli consisted of chips that, although electrically sound, contained visual anomalies. The first experiment found that the modal duration of eye fixations of trained inspectors was about 200 msec. The most accurate inspectors made the fewest eye fixations and were the fastest. The second experiment evaluated the performance of inspectors at one of the many sequential stages of chip inspection and found that 23% of the chips containing anomalies were accepted, whereas only 2% of the chips without anomalies were rejected. When the same chip was judged more than once by an individual inspector the consistency of her judgment was very high whereas the consistency between inspectors was somewhat less. The third experiment showed that variation by a factor of six in inspection speed led to variation of less than a factor of two in inspection accuracy. The fourth experiment showed that inspection via a ground glass screen is only a little worse than the usual method of looking through a binocular microscope. This was true even though the inspectors had no previous experience with the screen.
本文描述了一系列四项实验的结果，这些实验旨在总体上理解和改进视觉检查，特别是针对小型集成电路（即“芯片”）的视觉检查。实验刺激物是一些芯片，尽管它们在电气性能上完好，但存在视觉异常。第一个实验发现，训练有素的检查员的眼睛注视的模态持续时间约为200毫秒。最准确的检查员眼睛注视次数最少，速度也最快。第二个实验评估了检查员在芯片检查众多连续阶段之一的表现，发现含有异常的芯片中有23%被接受，而没有异常的芯片中只有2%被拒收。当单个检查员对同一芯片进行多次判断时，她的判断一致性非常高，而不同检查员之间的一致性则略低。第三个实验表明，检查速度变化六倍只会导致检查准确性的变化小于两倍。第四个实验表明，通过毛玻璃屏幕进行检查仅比通过双目显微镜观察的常用方法稍差。即使检查员此前没有使用该屏幕的经验，情况也是如此。
