| This research postulates that windows substitute for and aid associative memory in software development and support online tools. Prior windows research asserts that windows aid or relieve short term working memory. The software engineer relies upon long term associative memory developed in work experiences in considering all software structures related to problems when fixing them. The software quality engineer, not having done the development, has no equivalent long term associative memory and finds it hard to check changes. To correctly make approvals, software quality assurance must understand the interrelationships between modules, and must rely upon documentation as a substitute for the developers' associative memories. Windows can simultaneously display all document files related to a change, aiding associative memory and freeing other cognitive resources for the work at hand.; A cognitive systems engineering model of complexity is used as a human ergonomics factor in an abstract expression of the utility of a software quality assurance human-computer interface related to complexity and three information science attributes of the visual display tube. Viewability--an originally identified information technology construct--is isolated from accessibility, which is well documented. Optimum number of windows is also considered an information technology construct dependent upon complexity. This research asserts that viewability and accessibility are fundamental constructs of the visual human-computer interface, and that optimal number of windows is a component of accessibility. The model includes concepts of Rooms and panes.; A complete, 3 x 3 factorial, partially counterbalanced, block design, with three replications of viewability and most productivity measures, tests some hypotheses resulting from propositions that multiple windows increase interface utility and software quality assurance productivity. Maximum number of windows allowed operationalizes viewability, and task complexity implements complexity. The subject pool size was 24, and up to 648 observations yielded p values much less than the significance criterion of.05 with no interactions, indicating viewability depends upon number of windows allowed, is independent of complexity and does not increase productivity. Optimal number of windows was shown to depend only upon complexity, as were productivity measures of efficiency, effectiveness and agent satisfaction. |
