Premium
Must computer courseware evaluation be totally subjective?
Author(s) -
Micceri Theodore,
Pritchard William H,
Barrett Andrew J
Publication year - 1989
Publication title -
british journal of educational technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.79
H-Index - 95
eISSN - 1467-8535
pISSN - 0007-1013
DOI - 10.1111/j.1467-8535.1989.tb00272.x
Subject(s) - purchasing , computer science , presentation (obstetrics) , reliability (semiconductor) , interface (matter) , measure (data warehouse) , corporation , multimedia , operations management , engineering , medicine , power (physics) , physics , bubble , finance , quantum mechanics , database , maximum bubble pressure method , parallel computing , economics , radiology
In today's complex computer training market, flashy presentations frequently prove the most important purchasing element while instructional design and content take a back seat to form. Courseware purchasers obviously require an objective and generic measure of computer assisted learning (CAL) courseware. Since any such measure must include a human evaluator this is a difficult objective to attain. Funded by a grant from the Westinghouse Corporation, the Center for Interactive Technologies, Applications and Research (CITAR) at the University of South Florida attempted to break complex concepts such as instruction, management and user interface into component pieces small enough for objective evaluation. A study involving ten CAL packages produced agreements among an average of seven of eight raters and reliabilities around. 70, while an investigation of traditional perceptual evaluation procedures by the same raters on the same courseware produced far lower estimates for both agreement and reliability. Seeking reasons for these differences, a detailed investigation of courseware summaries showed that evaluators tend to pay more attention to the presentation than to the instructional aspects of the courseware. This phenomenon apparently has less influence on the low inference CITAR evaluation model which produces relatively consistent evaluations of CAL courseware.