Another undated WLC Technical publication, authored by JVM, 28-pages in length, that discusses accuracy levels, the types of errors that may occur in any such system. If JVM memory is correct, this document is based in large part on MRC experience in processing student tests with OMR technology, and furthermore, a series of simulated ballot-scanning tests conducted at the MRC Iowa City facility, prior to actual usage of the equipment in 'live' elections. Written as a 'primer' for users of the WLC Ballot-Scanning system. The 'conjectures' contained therein regarding accuracy were highly appropriate, and in the final analysis, attention to these details resulted in highly accurate processing and tallying of actual election-day ballots in the mid-1970's. The Archival DVD Media disc Folder: 1976_circa_BALLOT-SCANNING ACCURACY REPORT, contains 28 JPEG-Image files of the Original Report [not provided] and a 14-page JVM MS Word File transcribed copy: 1976_ca_BALLOT SCANNING ACCURACY REPORT_JVM.doc with a hard-copy printout attached