Human-computer interaction

CARING Newsletter, Fall, 2001 by Marilyn Herridge

This article will provide an overview of the variety of subjects to be considered under the umbrella of Human-Computer Interaction. As a recent graduate of the University of Maryland at Baltimore, Nursing Informatics Program, I had the privilege of studying with Dr. Norcio, Professor of Information Systems, from the University of Maryland Baltimore County. Dr. Norcio directs the User System Environment Research Laboratory, has been the Co-Director with Dr. Marion J. Ball of the Pan- American Health Organization/World Health Organization, is a computer scientist at the Artificial Intelligence Center of the Naval Research Laboratory and specializes in theoretical and applied areas of software and systems design, intelligent user interfaces and health informatics. Some of the facets of this subject to be considered include supportive theories and such topics as design methodologies, system usability, evaluation methodologies, user modeling, task modeling and natural language interfaces. More familiar topics include graphical user interfaces, virtual reality and multimedia systems, web interfaces, adaptive interfaces and intelligent agents.

The study of human-computer interaction (HCI) is based on a variety of theories that include the "Technology Acceptance Model" which contends that the design choices we make influence user acceptance. This model also includes the concepts that the perceived usefulness and perceived ease of use influence technology acceptance. The "Behavioral Model of Rational Choice" theory outlines the concepts involved in decisionmaking such as available alternatives, value or payoffs, outcomes expected and the likelihood of these probabilities occurring with each choice.

Theories of human problem-solving and theories of cognition include the natural approaches to the familiar process of problem-solving that allow us to separate the problem into memorable parts (chunking), analyze the known components and variables using known solution patterns, seek alternative solution methods and store learned techniques for future reference. Design rules, based on theories of human error are also critical to the success and acceptance of systems. Most errors are system induced--based on design flaws, not on user inability. By appreciating such natural errors, categorized as intentions, mistakes and slips, the designers can become sensitive and responsive to the needs of users.

Design methodologies are the principles of HCI used in the design process. Aspects such as an accurate problem definition, feasibility, cost-benefit analysis, project management methods, customer input, and the use of computer aided software engineering (CASE) tools in the development phase are among the necessary considerations that help to reduce the risk of project failure.

System usability allows the technology to recede into the background and allows the user to intuitively do their work. Usability must originate in the design phase and accommodate user input, testing and necessary refinements of the interface. Usability principles address speed, accuracy and satisfaction. Although functionality is a requirement of good software design, it does not guarantee usability. Factors to be considered include design consistency, efficiency, ease of use, format design (characteristics such as navigation, highlighting, color, and graphics), security, flexibility, and customization.

Evaluation methodologies include measurements of usability such as time to learn, speed of performance, error rates, retention over time and subjective satisfaction. These methodologies can be applied during the pilot phase (formative) or later in the life-cycle (summative).

User modeling incorporates the use of agents in software design that help us by suggesting alternatives, identifying new issues and evaluating differences of the decisions we make. While the user should have the feeling of maintaining control, agents provide built-in safeguards and feedback. The agents should blend transparently into the user's work environment.

Task modeling and task allocation techniques are employed in the early stages of the software development life-cycle. They drive the prototyping of user interfaces by predicting human behavior at the interface such as the predicted time to complete a task and errors that might be expected. Task allocation divides the labor associated with each task, strengthening efficiency and improved outcomes in system design.

Natural language and spoken language interfaces deal with the inconsistencies and variabilities associated with language. Design successes include: spell checkers, grammar checkers, multi-lingual dictionaries, search engines, translation software, and speech recognition.

Graphical user interfaces (GUI) allow the user to interact with the computer using graphics and pictures rather than entering typed text. The use of a GUI eliminates the need to memorize complicated commands by replacing such actions with icons, button sliders and menus. The GUI can provide for direct querying of underlying databases (direct manipulation queries).