Extract from the Proposal to the NSF program in Course, Curriculum and Laboratory Improvement (CCLI) Program

Due to burgeoning employment opportunities, courses in Web technology have become popular with Computer Science (CS) majors. New York State data indicates that there are over 3500 Web site development companies in the New York City area alone [NY State 1999] and an informal survey of SIGCSE members suggests that the demand for Web developers is nationwide [McCracken 2000].

CS majors have much to offer such employers: they know programming and understand networks. However, employers are looking for additional background in user-centered design, which is crucial to the success of any commercial Web site [Binstock 1999], [Roland 2000]. Of the 260 job announcements for Web developers sent to one of the PIs last year, over 97% mentioned knowledge of user-centered design in the qualifications [Wolfe 2000].

Studies have shown [Tedeschi 1999] that successful Web sites use a user-centered design approach. This approach is different from anything most CS majors now experience in the classroom. Most CS majors receive no exposure to user-centered design. Their mindset is geared to a systems development perspective and they have never considered the needs of users. They are unaware of the design aspects of Web site development and have no experience or education to help them organize a Web site from the perspective of its users.

Attempting to learn user-centered design by taking psychology or art courses does CS majors little good. While psychology courses do discuss cognitive limitations and protocols for controlled experiments, they are far too general to cover the specifics of user testing user interfaces [Hix 1993]. The courses in graphic design offered by art departments emphasize print media, but designs that are effective in print often fail as Web pages [Williams 1998]. Further, they do not address other areas of user-centered design such as content organization.

A much better approach is for CS departments to offer courses in user-centered design. Human-Computer Interaction (HCI), albeit a major category mentioned in ACM/IEEE-CS Curriculum 91 [ACM 1991] and Curriculum 2001 [ACM 2000], is still not taught in many CS departments. According to a recent survey of SIGCSE members [McCracken 1999] only half of all responding faculty said that their department offered any course in user-centered design or HCI, and only about 10% taught the course regularly. This is due to its relatively recent appearance as a discipline and to the fact that few faculty have any background in teaching the topic.

Knowledge of HCI is useful to all CS graduates, regardless of their career choice. The popularity of Web technology among CS students presents an opportunity to it to an audience that might otherwise be reluctant to engage in a methodology that is unfamiliar to them. Using Web development as the technological carrot to draw students into the course, instructors have the opportunity to discuss the principles and methodologies of HCI. The Web is an excellent vehicle for studying and practicing user-centered design. It is generally easy to build testable prototypes quickly. Because Web sites share many properties with other interfaces, lessons learned will translate well into other areas of software development.

Opportunity is always accompanied by challenges. For instructors, teaching a new course in an unfamiliar area is daunting enough; it may be deemed impossible when there is a lack of supporting materials such as sample syllabi, example homeworks or a suitable text. The Curricula for Human-Computer Interaction [Hewett 1992] forms a solid framework for developing HCI courses and cites a number of excellent texts, but because of their early publication date (pre-1992), they do not contain many specifics for using Web development as a vehicle for teaching HCI.

The PIs propose the design and test teaching of a curriculum that uses Web development as a vehicle for teaching user-centered design to CS majors. Since there are no available materials suitable for the course, the PIs propose to create them. Because the teaching techniques for this course are unfamiliar to most CS faculty, the supporting materials will include an extensive instructor’s manual.

The course will be tested at the proposers’ schools, and instructors at two other schools have agreed to test-teach the course and materials. Faculty workshops will extend the dissemination in a direct way, familiarizing instructors with the pedagogy and materials.

The goals of the project are:

The assessment plan for these goals appears in Section 4D, on the evaluation plan.