COMPUTER SELF-EFFICACY - Human-Computer Interaction and Management Information Systems: Foundations - page 218

manipulation for one software package (spreadsheet) but not another (word processing). Yi and
Davis (2003) have explored the four sub-processes of observational learning (Bandura, 1997),
and found that observational learning, which consisted of attention, retention, production, and
motivation dimensions, significantly influenced specific post-training computer self-efficacy.
Verbal Persuasion
Verbal persuasion refers to the use of suggestive persuasion by others to persuade an individual that
one has a given level of capabilities (Bandura, 1997). Verbal persuasion can increase SE when pos-
itive, realistic persuasive information is conveyed, which in turn results in greater effort and per-
sistence than if individuals have self-doubts when faced with difficulties. Individuals providing the
persuasive feedback needed to possess the evaluative competencies to perform actions/tasks (Bandura,
1986), and the individuals must be viewed as credible and believable (Bandura, 1977a, 1977b). As
suggested by Bandura (1978, 1986), verbal persuasion can increase one's self-efficacy, but is rarely
enough to create enduring high levels of efficacy beliefs. The greatest benefit of verbal persuasion
occurs in situations where individuals believe that they can produce effects through their behaviors
(Bandura, 1997).
Verbal persuasion in the IS literature (Table 11.A2) has been either measured directly or manip-
ulated in several studies. In studies using direct measures, verbal persuasion in combination with
vicarious experience, when used to manipulate general computer self-efficacy, resulted in an increase
in individuals' judgments of their capabilities to use computer technology (Hill et al., 1987; Smith,
1994) and the effects of verbal persuasion on GCSE held in another cross-sectional study (Compeau
and Higgins, 1995b), but not in a longitudinal study of the same subjects (Compeau et al., 1999;
Martocchio, 1994). Similar positive relationships between verbal persuasion and CSE have been
found in experimental studies conducted in a training context (see Table 11.A2 for a list of studies
in this area).
One's self-efficacy, according to Bandura (1997), can be influenced by the manner in which per-
formance feedback is structured or framed. Attributional or evaluative feedback can be communi-
cated in ways that enhance or undermine an individual's SE (Bandura, 1997). For example, Schunk
and his colleagues (e.g., Schunk, 1982, 1983, 1984; Schunk and Cox, 1986; Schunk and Rice, 1986)
investigated the effects of attributional feedback on SE judgments. They reported that attributional
feedback highlighting effort and ability as reasons for success enhanced children's mathematical
and reading SE judgments. From an IS perspective, Martocchio and Dulebohn (1994) reported that
trainees who received feedback attributed to factors within their control, a causal dimension of attri-
butions or causes of one's spreadsheet performance, demonstrated higher levels of SE, whereas
trainees who received feedback attributed to factors outside their control had lower levels of SE.
Physiological States
According to Bandura (1997), physiological and affective states have an influence on self-efficacy
judgments. The cognitive processing of emotional states is likely influenced by a variety of factors
such as the level of action, appraisal of the sources of the arousal, prior experiences on how the
arousal affects one's performance levels, and the conditions under which the arousal is elicited
(Bandura, 1986). The optimal level of the emotional state is determined by the nature of the task
plus the causal inferences concerning the arousal.
Certain negative emotional reactions such as fear, anxiety, and tenseness, or stressful situa-
tions, can lower computer self-efficacy. However, positive emotional states, such as excitement or