Computer-based objective interactive numismatic system
Three overreaching findings emerged from the details of this research investigation: (1) The Internet makes a feasible alternative to the palpable examination of rare coins for the purpose of grading: human numismatists accepted the validly of determinations they reached on the basis of images delivered on Web pages, and an automated grading system successfully operated with images of the same resolution as the delivered images. (2) The aggregated interagreement on coin quality among human graders working individually over the Internet was substantially increased when they were furnished with a machine-generated reference point. (3) The development of automated systems for grading coins that work with effectiveness of human experts appears to be a technical possibility based on the successful prototype described herein. ^ Taken collectively, these findings suggest that coin grading may be done more economically and in a timelier manner than it is done now through the mailing of coins to professional services. ^ There are two empirical thrusts. One examines the use of machine grades supplied as a grading aid to experienced human graders working over the Internet. The goal was to gain insight into the process of remote coin grading over the Internet. The presence of accurate computer-generated grades improved the human performance. The validity of the grading process was substantiated by significant test-retest reliability on repeated gradings of the same coins. ^ The other thrust explores the effectiveness of an automated system yielding expert grades on the 70-point Sheldon scale for Lincoln cents. The system matches a coin's histogram of red, green, and blue sub pixel frequencies against histograms in a database of coins rated by human experts. While there is little construct validity for this technique, data confirms that it works well. ^
Bassett, Richard A, "Computer-based objective interactive numismatic system" (2003). ETD Collection for Pace University. AAI3106821.
Remote User: Click Here to Login (must have Pace University remote login ID and password. Once logged in, click on the View More link above)