Where does computer self-efficacy collide?

A judgement of one's capability to use a computer to accomplish a task, rather than skill at any single package or command. Compeau and Higgins built the measure in 1995 by asking whether respondents could complete a job with an unfamiliar software package under ten decreasing levels of support, from being shown how first to having nothing but the built-in help, and kept Bandura's structure of magnitude, strength, and generalizability. Their 1999 longitudinal study of 394 end users traced the nomological net: computer self-efficacy fed outcome expectations, affect, and anxiety, and shaped computer use a year later. This is the referent construct for the map, the version of the belief that carries the name in information systems research.[2][3]

Unit

A user's belief about using a computer to complete a task

Recovery means

Not applicable

Measured by

Ten confidence items under decreasing assistance; use, affect, and anxiety as criteria

The conviction that one can execute the behaviour required to produce a given outcome, which Bandura set out in 1977 as distinct from the expectancy that the behaviour will in fact produce it. Efficacy beliefs are formed from mastery experience, vicarious experience, verbal persuasion, and physiological state, and they govern whether a person tries, how long they persist, and how they feel while doing it. Bandura held that efficacy is domain- and task-specific and should be measured against graded tasks in a named situation. Computer self-efficacy is this construct pointed at the machine; the site treats the wider family, including the general trait Bandura disowned, on its own page at self-efficacy.[1]

Unit

A person's belief about a specific task in a specific situation

Recovery means

Not applicable, efficacy is built rather than recovered

Measured by

Graded task lists rated for confidence at each level

Inside the technology acceptance model, computer self-efficacy is not the outcome of interest but an input to it. Venkatesh's 2000 anchoring-and-adjustment model cast self-efficacy as an internal-control anchor: before a user has touched a new system, their standing belief about their capability with computers sets an early perception of how easy the system will be, and that perception adjusts as hands-on experience accumulates. Venkatesh and Bala carried the anchor into TAM3 in 2008 as one determinant of perceived ease of use among several. The belief is the same judgement Compeau and Higgins measured, but its job in the theory is to explain ease of use rather than to predict system use directly.[4][5]

Unit

A user's control belief that anchors perceived ease of use

Recovery means

Not applicable

Measured by

Computer self-efficacy items as an antecedent of perceived ease of use