As a subfield of the academic field of management information systems (MIS), human computer interaction has traditionally been concerned with the process and outcomes of users interacting with computers to accomplish organizational tasks. The principal factors are the computer, the task, andmost importantthe physical, cognitive, and affective aspects of human users. The organizational context, which is a unique assumption of MIS, influences the theoretical basis for research into how these factors interact. Other fields that address HCI, such as computer science, human factors, information science, and cognitive psychology, do not always stress the organizational and workplace contexts or the organizational and managerial tasks.
Furthermore, the scope of HCI research in MIS has been comparatively wide. For instance, in the 1970s, when many users did not interact directly with mainframes but used printed reports generated by computer operators, HCI studies, such as studies of satisfaction , concentrated on the end userthe person who used the information rather than the people who operated the computer. Later, in the `80s, Jane Carey initiated a series of symposia called Human Factors in MIS , which included not only studies of users, but also of developers. Anything in the interplay between human agents, computer systems, and organizational tasks that made a difference to the business seemed to interest the field. But as we demonstrate below, much of the research was to be conducted in the tradition of social sciences that values theory and empirical-research methods.
Our review is limited to research that explicitly addresses or implicitly assumes experiences that include all three components: human, computer, and task. Very little research in MIS has concentrated on interactions between human and computer regardless of task, such as finding the universally best combinations of color, a topic commonly found in human factors. In most studies of attitudes and satisfaction in MIS, task is not articulated in the research model but is assumed to be part of a user's purpose for interacting. Moreover, as the context is usually organizational, the tasks to be accomplished are usually organizational tasks, such as data entry or business transactions (starting in the '70s), decision making and office work (starting in the '70s), communication and collaboration (starting in the '80s), and online navigation and purchasing (the past decade). Due perhaps to the centrality of the organizational context and task, there has been very little research on physical aspects of design, such as the ergonomic design of keyboards, because these devices are usually designed for a wide range of tasks and treated as a given in MIS, although they are popular research topics in fields such as human factors and industrial engineering.
The past four decades have been associated with substantial developments in information technology: changing hardware platforms (mainframes, mini-computers, micro computers, networks/Internet, and mobile) and the introduction of major applications (word processing, spreadsheet, email, Internet browsers, IM, enterprise-wide systems, etc.). Figure 1 plots these developments in the first and second rows; the third row describes the major tasks associated with these applications; the bottom row identifies frequently studied topics. Generally speaking, the computer platforms, applications, and tasks dictated, at least initially, the research topics, and usually with a few years' time lag.
New technologies often grab the field's attention; soon after they appear, there is a surge in descriptive studies of how they are used. For instance, mainframes and transaction processing systems triggered research on errors and error messages, effective data entry, and report formatting (see Figure 1). Mini and micro computers created interactive environments that supported decision making, which in turn created interest in the impact of interactive input-output design on decision effectiveness, dramatically reducing interest in printed reports. With the expanding usage of networks and the Internet, communication support and collaboration became major issues, and with the growing interest in e-commerce, website usability and trust have become popular topics. In parallel, we can see a shift from primarily individual-level research to a more balanced portfolio, with group-level research too. Among current rapidly growing topics are attitudes and use, affect, and trust (see a thorough review of topics studied in the `90s ). To sum up, Figure 1 plots illustrative research topics in relation to the technology and task of each period. The research largely ignored physical aspects and focused on cognitive aspects. However, research on affective aspects is quickly growing in popularity, and with the application of wearables and miniature computing devices in the workplace, we may witness a new interest in physical aspects, too.
The centrality of task in MIS showed up early in a series of cognitive and behavioral studies that examined how alternative designs impact effectiveness, which came to be known as the Minnesota Experiments . First, the task was usually managerial and mainly decision making; tasks and their effectiveness were measured by decision performance. Second, designs included both the format of presentation and when, what, or how much information was presented. Thus, information overload was examined as an HCI issue in supporting decision making. The research demonstrated the importance of user-computer-decision interaction, which affects users' performance as well as their assessment of the information system. The researchers concluded that while computer characteristics (such as CRT output) affected managerial work significantly, the impact of user characteristics was unclear.
Although developments in IT pushed researchers to focus on new technologies, the field (through journal editors, dissertation supervisors, etc.) has consistently demanded more theory-based research and more sophisticated measurements. The MIS field advocates theory development as essential for advancing knowledge, believing that the discovery of general laws and when to apply them should result in more flexible and more effective designs. The case of research on individual characteristics, particularly cognitive style, demonstrates this trend well. Cognitive styles are consistent patterns of organizing and processing information, and would therefore appear to be relevant to HCI design. In response to the emphasis on decision support and the ease of providing business graphics, there was a surge of empirical studies about the effect of cognitive styles on the way information should be presented, particularly whether to use tables or graphs.
In a controversial 1983 article entitled "Much Ado About Nothing," Huber  slashed into the research on cognitive style, claiming it had not in the past led to any useful conclusions and was unlikely to do so in the future because it was based on inadequately developed theory and employed faulty measures and research designs. Research on individual differences has since all but stopped (a few recent exceptions may signal a revival). But the call for a better theoretical basis has been echoed continuously in other research topics such as satisfaction  and decision support . Interestingly, research on group decision support was relatively rich with theory in comparison with individual decision support, and consequently sustained longer . Parallel calls for better measurement appeared from time to time, particularly for field surveys and laboratory experiments (e.g., ). In response, several general theories or models of HCI emerged that influenced many subsequent studies.
The most cited model is TAM: the Technology Acceptance Model  and its extensions. TAM explains the user's decision to adopt a system with implicit reference to task and technology. A user's intention to use a system is determined by its perceived usefulness and ease of use, which may both be influenced by situational, organizational, and personal factors. In response to the ubiquity of computing and the blur between work and leisure computing, perceived enjoyment recently became another factor affecting technology acceptance.
In addition to studying technology acceptance, several models attempted to explain interaction by considering the fit between user, computer, and task. Cognitive fit theory  explains how fit (or lack of fit) between task requirements, user representation, and computer presentation affect a user's behavior. For example, presenting textual information for a spatial task reduces task performance. Similar frameworks address the group level . Additionally, researchers borrowed theories from fields related to business and management in order to understand the particular task at hand and integrate it into specific human-computer interaction, such as theories of decision behavior, communication, information value, and consumer behavior . These theoretical advances have been critical in promoting HCI publications in MIS.
Another important factor in promoting HCI in MIS is its institutionalization within the international Association of Information Systems (AIS) after a long period of unorchestrated local initiatives. Table 1 shows some key stages. The University of Minnesota, which was the first of several MIS centers, launched the individual-level Minnesota Experiments and later group decision support systems research. A series of symposia on Human Factors in MIS began in the mid-'80s . Fifteen years later, the AIS formed the special-interest group on HCI (http://sigs.aisnet.org/SIGHCI/), which became one of the largest and fastest-growing SIGs within AIS. Presently, it publishes a newsletter twice a year, holds annual workshops and conference tracks in all major AIS conferences (ICIS, AMCIS, ECIS, PACIS) and in HICSS, and sponsors special issues in major MIS and general HCI journals. The last year saw the publication of two edited research collections  and the first textbook .
The subfield of HCI in MIS has been growing for 30 years, with a recent surge of activity. Organizational task and context have been central to this research. It has concentrated on cognitive aspects but recently opened up to include affective aspects. Technological innovations trigger research initially; accumulated knowledge helps uncover more general laws and eventually inform the design and use of future technologies. I believe HCI research will continue to follow these trends, capitalizing on organizational and managerial theories to better understand the link between user, computer, and task. I also believe that the idea of integrating theories of organizational task into HCI will be practiced, with appropriate theories, in other applied fields that are characterized by other types of tasks, as is already evident in education, for example. As we expand the diversity of tasks and contexts for which we study HCI, it may be time to study human computer interaction at two inter-related levels: the human-computer-task interaction that explicitly considers what the higher-level task users' wish to accomplish and the lower level that considers how users interact with computers to enable the higher level . And this may result in more cases of human-computer-task interaction developing as subfields of applied research fields.
Many thanks to Jonathan Grudin and to Izak Benbasat, Jane Carey, Phillip Ein-Dor, Dennis Galletta, and Ping Zhang for their suggestions and feedback.
9. Newsted P.R., W.D. Salisbury, Todd P., and R.W. Zmud. "Measurement issues in the study of human factors in MIS," in Human Factors in Information Systems: The Relationship Between User Interface Design and Human Performance, edited by Jane M. Carey. Greenwich, Connecticut: Ablex Publishing Corporation, 1997.
13. See, respectively: Todd, P., and I. Benbasat. "Evaluating the Impact of DSS, Cognitive Effort and Incentives on Strategy Selection." Information Systems Research 10, no. 4 (1999): 356-374; Srinivasan A., and Irwin G. "Communicating the Message: Translating Tasks Into Queries in a Database Context." IEEE Transactions on Professional Communication 49, no. 2 (2006): 145-151; Rafaeli, S., and D.R. Raban. "Experimental Investigation of the Subjective Value of Information in Trading." Journal of the Association for Information Systems 4, no. 5 (2003): 119-139; Koufaris, M. "Applying the Technology Acceptance Model and Flow Theory to Online Consumer Behavior." Information Systems Research 13, no. 2 (2002): 205-223.
14. Human-Computer Interaction and Management Information SystemsFoundations. Edited by P. Zhang and D. Galletta. Armonk, NY: M. E. Sharpe, Inc., 2006; Human-Computer Interaction and Management Information SystemsApplications. Edited by P. Zhang and D. Galletta. Armonk, NY: M. E. Sharpe, Inc., 2006.
Tel Aviv University
Assisted by Orly Amramy
ABOUT THE AUTHOR
Dov Te'eni is professor of information systems at Tel-Aviv University. He studies several related areas of information systems including human computer interaction, computer support for communication, knowledge management, and systems design. His research usually combines model building, laboratory experiments, and development of prototypes like Spider and kMail. Dov is coauthor of a recently published book on human computer interaction for organizations.
©2007 ACM 1072-5220/07/0700 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2007 ACM, Inc.