This Technical Brief covers the testing of the DIgitally Controlled Analog Discrimination (DICAD) system developed by MRC Engineering to convert the students' pencil-marked answers on their test sheets (during scanning) into a 4-level digital binary code, defined as High, Medium, Low, or Zero, thus representing an accurate comparison of relative marking levels for separating intended marks from erasures, smudges, and stray marks or doodles. Early MRC hardware designs sampled only one small area, ideally in the center region of the marking 'bubble' or target. In later MRC embodiments of the DICAD scheme, multiple CCD-pixel sample points were taken within the area of the student's marking 'bubble' (e.g., a small circle, oval, or box wherein the student placed his/her answer mark) and each sample was converted into the DICAD code, then 'added' to all the other sample points taken within the bubble. Thus, a wide range of digital codes was generated for computer analysis of each marking position, from "000" up to "255" (e.g., an 8-bit byte per bubble). This refinement led to an extremely accurate discrimination capability between the student's intended marks and ones that should be ignored, and also effectively dealt with the vexing but common situation of student "A's" average marking level or density being significantly different from student "B's" average marking level: due to different pencils used, marking pressure, muscle strength, etc. Refer to McMillin U. S. Patents No. 3,486,040, No. 3,524, 048, No. 3,692,982, No. 3,820,068, and McMillin/Schroeder U. S. Patent No. 4,300,123 for a review of the evolution of this important MRC milestone in constantly improving the accuracy of scored answer sheets at the Iowa City facility, and later in other worldwide installations of MRC scanning systems