HUMAN FACTORS, CHI, AND MIS - Human-Computer Interaction and Management Information Systems: Foundations - page 393

knowledge acquisition, requirements engineering, and a broad range of human factors topics. MIS
HCI research appears in the middle row, with the bottom row representing CHI and its predecessors
that focused on discretionary use—support for engineers and programmers, and visions of writers
and prototype builders.
Books titles are italicized ( POET represents Norman's 1988 Psychology of Everyday Things ).
Martin's 1973 topic devoted to designing for computer operation and data entry is in the top row,
whereas he is listed among the visionaries for the topic's first chapter. Card, Moran, and Newell's
1983 topic is in the top row, despite the authors' prominence in CHI and their disassociation from
human factors, because their models of expert performance were intended to advance human fac-
tors, staying close to the physical interface and addressing tasks such as non-discretionary key-
board use by telephone operators. As described below, when HF&E became receptive to cognitive
theory it embraced this modeling research.
1985-1995: GRAPHICAL USER INTERFACES AND HCI BEFORE THE WEB
The graphical user interface or GUI was envisioned and features were prototyped in the 1960s.
Johnson et al. (1989) describe GUI development at PARC in the 1970s. But only when the 1984
Apple Macintosh survived a rocky first year did the GUI or direct manipulation interface transform
HCI research. The Mac did not fare well in the large enterprises that were initially targeted, but it
appealed to graphic artists and other discretionary users. As a result, the impact registered first in
CHI. Even there, a prominent theoretical analysis by Hutchins, Hollan, and Norman (1986) con-
cluded that it was uncertain how GUIs would fare. They identified benefits for new and casual
users, but also noted possible advantages to interfaces based on commands and function keys for
skilled users. Mouse-wary corporations, concerned with skilled use, did not embrace GUIs until
Windows 3.0 succeeded in the early 1990, and even then not for tasks such as data entry.
Human Factors and Ergonomics
Government agencies, such as those responsible for the military, social security, census, aviation,
and taxation, and private industries, including the financial, insurance, aviation, medical, and
telecommunications sectors, continued to rely on systems with captive users. Funding at U.S.
research agencies such as NSF and DARPA focused on non-discretionary uses of technology.
This included massive investment in speech recognition and natural language processing, tech-
nologies avoided by users who have a choice, who are not constrained by disability, occupied
hands, or the need to navigate a phone answering system.
In 1986, Smith and Mosier published the last in a series of interface design guidelines sponsored
by the U.S. Air Force. Guidelines were a tool for writing contracts: An interface could be required
to adhere to specific guidelines. Smith and Mosier mentioned GUIs, but their 944 guidelines did
not address them. The vastly increased GUI interaction capabilities—pop-up and pull-down menus,
windows, mouse button assignments, color, sound, animation—rendered a comprehensive set of
design guidelines unattainable. Process-oriented approaches were needed for specifying interac-
tion design.
A HFS Computer Systems Technical Group study of the potential impact of CHI found an
unexpectedly small overlap of focus. CSTG hewed more closely to its original focus on hardware
(packaging, input devices, displays) and drew psychologists working in the milieu of military and
government contracting, whereas CHI focused on innovative commercial software and its devel-
opment (Richard Pew, personal communication).