Webb, S., and MacMillian, J. Cognitive bias in software engineering. Communications of the ACM, 38(6), June 1995, pp. 57--63.  Home/Search   Document Not in Database   Summary   ACM   TOC   Related Articles   Check

....to model. And without a model for biases in software development it would be difficult to trace how the reengineering provided by cognitive support would impact them. Fortunately there may be some promise in simply building a list of cognitive biases in software development (see Stacy et al. [691] for a start) After building the list one might be able to employ empirical legwork to determine how cognitive support affects them. In this, at least, CoSTH makes such an empirical endeavor possible by providing a list of experimental variables (i.e. the support types) to consider. Freed et ....

Webb, S., and MacMillian, J. Cognitive bias in software engineering. Communications of the ACM, 38(6), June 1995, pp. 57--63.

....knowledge in software engineering about roles, procedures, standards, software process, just to mention a few. Without an adequate field study, it is difficult to predict what knowledge software designers are missing (see conclusions) 4.5. Cognitive Bias According to the literature in cognition [31, 26], people s intuitive inference and probability judgments do not strictly conform to the laws of logic or mathematics. People rely on a limited number of heuristic principles which reduce the complex task of assessing probabilities and predicting values to simple judgmental operations. In general, ....

....whose instances are less retrievable. Then, vivid, compelling or otherwise interesting instances can be overweigheted in assessing the 2 For example, UML[1] defines 9 types: use case, class, object, sequence, collaboration, statechart, activity, component and deployment. 3 Stacy and Macmillian[26] call bias what Tversky ## ###[31] call heuristics. frequency or probability of such events because they are so easily remembered. This explains why there are wrong perceptions of contributions in teams[19] it is easier to remember our own contributions than the contributions of the others. ....

W. Stacy and J. Macmillian. Cognitive Bias in Software Engineering. Communications of the ACM, 38(6):57--63, June 1995.

....should support the construction of several linked views representing a variety of cross referenced mental models. E6: Support the construction of multiple mental models Not only do mental models differ in context and level of abstraction, but they also differ from one maintainer to another [38]. Several mental models of a program may be presented visually using multiple views. Many of the tools already mentioned support multiple views of textual and graphical views (PLUM [19] Rigi [39] Whorf [22] Garden [40] VIFOR [21] CARE [23] SeeSlice [27] and Imagix [25] Example graphical ....

W. Stacy and J. MacMillian. Cognitive bias in software engineering. Communications of the ACM, 38(6):57--63, June 1995.

....should support the construction of several linked views representing a variety of cross referenced mental models. E6: Support the construction of multiple mental models Not only do mental models differ in context and level of abstraction, but they also differ from one maintainer to another [53]. Several mental models of a program may be presented visually using multiple views. Many of the tools already mentioned support multiple views of textual and graphical views [27, 29, 30, 31, 33, 35, 45, 54] Example graphical views show call graphs and variable usage diagrams. Example textual ....

W. Stacy and J. MacMillian. Cognitive bias in software engineering. Communications of the ACM, 38(6):57-- 63, June 1995.

....been found by type checking. The defect detection capabilities of testing methods [2, 8, 22] have received some attention; the corresponding psychological problems were also investigated [20] There is also a considerable literature about debugging, e.g. 7, 13, 16, 17] and its psychology, e.g. [17, 19]. However, the defects found by testing or debugging are those that already passed the type checks. So the results from these studies would be applicable here only if they focused on defects detectable by type checking which they do not. Several studies have compared the productivity effects ....

Webb Stacy and Jean MacMillian. Cognitive bias in software engineering. Communications of the ACM, 38(6):57--63, June 1995.

....should support the construction of several linked views representing a variety of cross referenced mental models. E6: Support the construction of multiple mental models Not only do mental models differ in context and level of abstraction, but they also differ from one maintainer to another [148]. Several mental models of a program may be presented visually using multiple views. Many of the tools already mentioned support multiple views of textual and graphical views [6, 14, 64, 86, 100, 120, 124, 125] Example graphical views show call graphs and variable usage diagrams. Example textual ....

W. Stacy and J. MacMillian. Cognitive bias in software engineering. Communications of the ACM, 38(6):57--63, June 1995.

....activity. The reasoning behind this is from experience in interviewing students about their contribution to projects, where often all students would claim they put in a greater effort than the others involved in the team. This observation has also been confirmed by research in the United States [6]. It was also hoped that having students note both start and end times would make them more likely to record the log during each activity, rather than at the end of the project. 2.2 Error measurements The other data that students collected on their projects was the number and type of errors ....

Stacy, Webb and Jean MacMillian. "Cognitive Bias in Software Engineering." Communications of the ACM, Vol. 38 Num. 6: ACM Press, June 1995, pp 57-63.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC