Up until the IBM 805 was introduced, test scoring was an entirely manual operation. Tests could be graded only as fast as a human could evaluate the marks on the page with his or her eyes. The speed of the IBM 805 was limited only by the operator's ability to insert sheets into the machine and record the scores. An experienced operator could record scores on answer sheets at the rate of about 800 sheets per hour or call the scores to a recording clerk at the rate of 1000 sheets per hour.
Resistance was the key
Rey Johnson wasn’t the only one trying to find a way to automatically read marks on a test answer sheet, but he was the first to find a practical way to do so. Because of the variability of the marks on the page, it was difficult to find a reliable way to read the mark. Johnson and other groups were looking at the electrical conductivity of a mark, but only Johnson realized that by introducing higher resistance into the sensing circuitry, he could read a wider range of correct marks.
Various IBM equipment could be used with the “mark sense” forms on which test answers were recorded. These were the IBM 513 and IBM 514 Reproducing Punches, the IBM 557 Alphabetic Interpreter and the IBM 519 Electric Document Originating Machine. Various combinations of these machines could be used to create punched-card data without using a keypunch machine. The pencil marks could be scanned and quickly turned into punches on the cards to automate record keeping for long-distance calls or meter readings.
From electrical to optical recognition
Although the IBM 805 was removed from marketing in 1963, the system was just the beginning of automated scoring technology that continues today. The 805 used electrographic technology, sensing the conductivity of the marks on the page. As the 805 was reaching end of life, IBM developed a scoring machine based on optical mark recognition (OMR), introducing the IBM 1230 optical mark scoring reader in 1962. The IBM 1230 could read and score 1200 test answer sheets per hour of unattended operation. OMR is still used today in scoring standardized tests and in products such as IBM FileNet ® Capture Advanced Document Recognition and IBM SPSS ® Data Collection Scan.