Is HCI Education Getting a Passing Grade from Industry?

URL: http://www.acm.org/~perlman/chi94edu.html

Table of Contents

THE PANELISTS

Gary Perlman
Computer Science, Ohio State University, USA, perlman@acm.org
Arye R. Ephrath
Usability Engineering, Bellcore, USA, arye@cc.bellcore.com
Thomas T. Hewett
Psychology, Drexel University, USA, hewett@duvm.ocs.drexel.edu
John Long
Ergonomics, Univ. College London, UK, j.long@ucl.ac.uk
S. Joy Mountford
Design Center, Apple Computer, Inc., USA, S.Joy@AppleLink.Apple.com
Jenny Preece
Computing, The Open University, UK, J.J.Preece@open.ac.uk

ABSTRACT

ACM SIGCHI is expending more and more effort on HCI education, funding a variety of projects, including a curriculum report, and IFIP working group, and a survey of programs. Is SIGCHI directing its resources wisely and effectively? Who are the consumers of HCI education, and who are their employers? What should be the direction of HCI education in the 90's? Panelists from industry and academia will "face off" to debate what industry wants, needs, and is getting from HCI education.

KEYWORDS: [K.3.2] Computer and Information Science Education, Curriculum; [K.6.1] Project and People Management, Staffing, Training. Human Factors, Management, Education

INTRODUCTION

ACM SIGCHI is expending more and more effort on HCI education. It has funded a variety of education projects, including a curriculum report [Hew92], IFIP working group 13.1, and a survey of programs [Perl93]. Is SIGCHI directing its resources wisely and effectively? Who are the consumers of HCI education, and who are their employers? What should be the direction of HCI education in the 90's?

The multiple skills needed to develop effective user interfaces make HCI education a demanding endeavor, but one that is critical for continuing progress in the field. Although most CHI attendees are not educators, all have been consumers of education, and many are consumers (employers or coworkers) of the product of education. CHI attendees form the market for students trained in HCI, and it is in the best interest of HCI for education and industry to communicate and cooperate.

Given these broad concerns about HCI education, the focus of this panel is how to meet the needs of industry, arguably the largest employer of students. In particular, we will not focus on PhD- level education, which while important, is not as critical to economic infrastructure as undergraduate and masters education.

PANELIST POSITIONS

The panelists come from: academia and industry, the USA and the United Kingdom, computer science, psychology, and engineering. They will discuss their views of the problems of meeting industry needs and what can be done about them.

Gary Perlman

When I began teaching user interface development to master's level software engineers at the Wang Institute, most of these experienced programmers had never seen a person using their software. My course evolved (regressed) to the point where the goal was to have the students evaluate their own design, if only to find out that their users found it unusable. At Ohio State, the situation is even worse, because the students do not have the experience of developing any real software, and they do not yet have the skills to bring their ideas to life. Most computer science departments view software engineering as beneath the dignity of the field, viewing projects (especially group projects) as a way to let weaker students get by. HCI is viewed with even greater disdain. The same is true in other departments, such as psychology, where human factors is viewed as too applied. One result is that I see constant complaints from both students and employers about inadequate preparation for the real world.

The answer is not to revamp education. Employers like the subjects that students learn in computer science and are not willing to forego any of them. But there should be more group- and project-oriented work integrated into the curriculum to reinforce both practical applications and theoretical implications.

Arye R. Ephrath

Fresh graduates need to hit the ground running in an industry environment. They don't, because at first they cannot do any useful work. They do not know how to run user tests, they cannot implement a graphical user interface, they do not know how to work in teams, they have no appreciation for the importance of making tradeoffs and no practice in making them.

Educators do provide their students with good grounding in the fundamentals, as well as in the methodology of HCI research. Above all, students learn in school how to learn, so they are trainable when they come to us. We would like to see them having more exposure in school to practical work: Iterative design of real user interfaces using real subjects; exposure to creating requirements and design in large scale software development environments, and learning how to share information with those who are doing the architecture and coding; and exposure to the basics of computer hardware and software design.

In most cases, Computer Science graduates and HCI folks collaborate for the first time in the work environment, and often it proves to be a struggle for both sides. Introducing them to each other in the simulated environment of the school might prove beneficial to both disciplines, as well as to industry as a whole.

Thomas T. Hewett

The fundamental problem for HCI education and practice is change. The world of computing technology changes rapidly, and will continue to do so. The learning of content done in a particular institutional context which assumes a clearly defined stopping point is not sufficient to prepare people to deal with rapid change. Rather, education and training must be viewed as on- going processes which continue throughout an individual's entire lifetime. Traditional approaches to the teaching-learning process need to be modified to stress an emphasis on critical reflection. It is reflection on process, on practice, and on content which enables learners to apply their knowledge to novel situations and to continue to learn in the future.

One step which might be taken is to create an operating "brokerage" arrangement for establishing relationships which would enable faculty, researchers, developers and designers to gain work experience in alternative professional settings (e.g., faculty would work on the industrial development of systems; researchers and developers from corporate labs would teach courses and benefit from the challenges which students can pose; industrial developers would work in research labs, etc.). Such opportunities for retooling, reskilling, and refreshing knowledge and skills would enable individual reflection on knowledge and practice, contribute to the growth of shared understanding critical to the field and foster a process of continuous learning.

John Long

'Good' students, successfully completing a 'good' HCI course, produce a passing grade from industry for HCI education. This grade is not awarded by industries recruiting 'less good' students, who have been 'less well' educated. Industry dissatisfaction has its origins in the state of HCI education, and of the HCI discipline.

HCI education should teach students to solve the design problem of users interacting with computers (and other users) to perform work effectively. Craft approaches rely on experiential knowledge and 'trial and error' design practices, and so are difficult to teach. Applied science approaches rely on scientific knowledge, and 'specify and implement' design practices, and so are difficult to integrate into the design process. Thus, students acquire insufficient craft skills, and their applied science skills are insufficiently integrated with design. HCI education can obtain a passing grade only by providing 'good' training to 'good' students.

Concerning the HCI discipline, industry may be dissatisfied because appropriate HCI knowledge and practices do not exist, and so even 'good and well educated' students may be unsatisfactory. For example, HCI has more evaluation methods than design methods. In addition, HCI substantive knowledge, e.g. guidelines and standards, has no associated guarantee. The situation will only ameliorate as HCI improves its knowledge and practices.

In summary, there are two origins for not awarding HCI education a passing grade - suboptimal training, and suboptimal HCI discipline knowledge and practices. 'Good' students with 'good' education minimise the problem. The opposite, unfortunately, is also true.

S. Joy Mountford

Successful education of Interaction/Interface Designers must be guided by two important principles; First, teaching design must incorporate an apprenticeship process; Second, educating designers must encourage interdisciplinary team activity. Such qualities are often difficult to reinforce and create from within any one university department. These needs also bring up concerns such as; 'Who / where are the great interface designers?' and 'How can we best evaluate team-based projects?' As the definition of interaction design broadens, the need to educate more interdisciplinary-trained team-oriented students increases.

I believe that the best solution to providing more meaningful education for interface designers is to create partnerships with industrial clients, who have experience both in solving real world problems and in team-based activities. This goal prompted me to establish the annual Interface Design project at Apple Computer. Selected universities are challenged to answer an interface design problem brief. Universities must work across departmental boundaries such as Computer Science, Psychology and Design, to form project teams to submit interactive software and hardware prototypes for industry critique. The results are very positive; professors formed new faculty relationships, students enjoyed the project and felt it added significant value to their education, and industry was happy to encourage a new generation of interaction designers. Apprenticeships in such interdisciplinary teams will facilitate the design of more usable and aesthetic products for everyday users.

Jenny Preece

My position statement is about providing relevant HCI training for people who are already employed in industry. People learn best when they see the relevance of what they are learning and when they are actively engaged in and can direct their own learning processes. This makes it difficult to provide a formula for HCI training in industry because: How do we proceed? We should give-up second guessing what students from industry need and make our courses sufficiently flexible for students to tailor them to their own specific needs. Longitudinal projects selected by the students themselves and based on their own work encourage students not just 'to see' but 'to experience' the relevance of HCI in a personal and meaningful way.

ACKNOWLEDGMENTS

In 1992, Michael Rees of Bond University invited the organizer to participate in an education panel at OZCHI'92; the complaints from the industrial attendees motivated the creation of this panel.

Companion Proceedings of the CHI'94 Conference on Human Factors in Computing. Copyright (c) 1994 by the Association for Computing Machinery, Inc. Copying without fees is permitted provided that the copies are not made or distributed for direct commercial advantage and credit to the source is given. To copy otherwise, or republish, requires a fee and/or specific permission.

BIBLIOGRAPHY

[Hew92]
Hewett, T. T. et al (1992) Curricula in HCI. ACM.
[Perl93] Perlman, G. & Gasen, J. (1993) The ACM SIGCHI HCI Education Survey. (poster submitted to CHI'94).

SUMMARY SLIDES FROM CHI'94

The Problem

  • Students (Graduates)?
    • have little practical experience / skills
    • can't (weren't allowed to) work in teams
  • Faculty?
    • probably do not have practical experience
  • Universities?
    • departments hinder multidisciplinarianism
    • elitist CS/psych view SE/HF as too applied
  • Industry?
    • doesn't know what they need
    • doesn't appreciate issues in HCI
  • Engineering Education?
    • design is difficult to teach
    • engineering grads require further training
  • Young/Changing HCI Field?
    • parts are ill-defined / of unknown value
    • computing technology is changing rapidly
    • knowledge and practices do not yet exist

The Solution

  • Design Projects
    • require that students make decisions
  • Motivation
    • provide realistic projects
    • have students select projects
    • have students evaluate their work
  • Interdisciplinary Teamwork
    • learn about complementary skills / fields
    • learn how to work in teams
  • University / Industry Partnership
    • students as apprentices
    • practitioners as mentors / liaisons
    • faculty gain experience / connections
  • Lifelong Learning
    • stress critical reflection