help fill in the gaps and blind spots of the team’s mental model of thetechnology. Pacey gets at this idea of adjusting the model through interaction: “…although ideasmay arise in all sorts of ways that may be described as intuitive or participatory, there is alwaysan obligation to translate them into more rigorous, often mathematical formulations, so thatothers may understand and check them, and explore their precise implications.”226. Bias in the Design ProcessLet us now explore the means by which bias may be introduced during each stage of the designprocess. Engineering design projects typically begin with a problem specification phase. Theconstraints, requirements, and specifications of the design are elicited from the customer
, and Inc. National Action Council for Minorities in Engineering. "Amicus Brief to the Supreme Court of the United States in Re: Nos. 02-241, 02-516, Barbara Grutter V. Lee Ballinger & Jennifer Gratz and Patrick Hamacher V. Lee Ballinger."Cambridge, MA: Massachusetts Institute of Technology, 2003. Available at http://www.umich.edu/%7Eurel/admissions/legal/gru_amicus-ussc/um/MITfinal-both.doc36. Denison University. "Improving The Economics Curriculum With Laboratory Experiments." Lessons Learned From FIPSE Projects II. Fund for the Improvement of Post-Secondary Education of the U.S. Department of Education. U.S. Department of Education, Washington, DC, 1993.37. Sullivan, W. G
the 2004 American Society for Engineering Education Annual Conference & Exposition Copyright 2004, American Society for Engineering EducationAn important observation is that our educational approach should not be “classical methods ormodern methods”, as there is a role for both, along with basic theory, principles, and equations,and other educational resources such as case studies and descriptive texts on general designconsiderations, project performance, and failure analyses. The challenge is how to makeoptimum use of this increasing rich mixture of educational resources to best prepare ourgraduates to effectively function in the computer/software intensive design environment so thatthey will consistently produce
to bothoperating systems on identical hardware specifications.Mail throughput is collected using the mail benchmark Postal (version 0.62), downloaded fromhttp://sourceforge.net/projects/postal/. Postal floods the specified server with SMTP messages ofrandom content and length, according to given parameters such as thread count, throughputthrottle, and connections. For the purpose of testing, Postal was run using unthrottledthroughput, a configuration of one message per connection, and variable thread counts.All performance tests are automated through configuration of the Cron Daemon in Linux andScheduled Tasks in Microsoft Server 2003. Each test spans 30 minutes, allowing for throughputand availability information to be collected. Server
(two or three days on a Sun enterprise 450) for 50 to 100 student submissions, thesearch window (range of possible relative file offsets) was constrained to as little as 100characters. Pre-filtering of the files, similar to the tokenization described above, was used toreduce the size of the files to be compared. However, restricting the search window reduced theinstances of plagiarism that could be detected.ResultsInitial testing was done using student source code samples from a simplified I2C bus interfacedesign project during spring 2003. I2C is an industry standard synchronous serial bus interfaceused in a wide range of consumer electronic products. Transformations were applied to thesource code samples in order to observe the behavior of