COMPUTER SELF-EFFICACY - Human-Computer Interaction and Management Information Systems: Foundations - page 208

Figure 11.1 Social Cognitive Theory

Behavior

Person:
Cognitive and Other
Personal Factors

Environment

CSE was influenced by a behavior modeling training program. In another study, the same authors
found that CSE was influenced by social influences in the organization (Compeau and Higgins,
1995b). Several authors have examined demographic and personality variables purported to influ-
ence CSE development (e.g., Henry and Stone, 1999; Thatcher and Perrewe, 2002).
At the same time, important questions about computer self-efficacy remain unanswered. Marakas
et al. (1998), following Bandura, argue for the need for more domain specific measurement of CSE.
Yet precisely where the domain should be defined is unclear. Is it a software package, a set of features
in a software package, a computing platform (e.g., the PC)? How does task domain interact with tech-
nology domain in the definition of CSE? These are difficult questions about the very definition of the
construct, which little research has addressed.
The purpose of this paper, then, is to review the state of research on computer self-efficacy in
the information systems literature, with particular attention to its conceptualization, its influence,
and its formation. We include a discussion of several research programs that we are undertaking
that begin to address gaps and opportunities uncovered in our review of the computer self-efficacy
literature. The paper concludes with an agenda for future research to further explore this impor-
tant individual cognition.

THE NATURE OF COMPUTER SELF-EFFICACY

Computer self-efficacy is an adaptation of the broader notion of self-efficacy, which itself is con-
ceptually situated within the broader context of social cognitive theory (Bandura, 1986). Within
his notion of triadic reciprocality (Figure 11.1), Bandura lays a foundation for understanding the
complexities of human functioning via the interactions that occur between the person (specifi-
cally his or her cognitions and emotions, including self-efficacy), his or her behaviors (including
learning and performance), and the environment in which he or she is engaged. Accordingly, each
of these three factors is seen as influencing the others while also being influenced by them. These
complex interactions are difficult to study, yet Bandura's theorizing also provides us with the
logic to tackle this complexity systematically. By theorizing and developing the causal sequence
of a phenomenon within a specific and limited time frame, researchers can select a shortened and
non-reciprocal causal chain to study—for example, that computer training design (an environ-
mental factor) influences computer self efficacy (an individual factor) which in turn influences
performance with computers (a behavior). The use of this strategy is implicit in much of the
research on computer self-efficacy. We use this broad conceptualization of triadic reciprocality
as an organizing framework throughout the paper to frame and illustrate the linkages between