How to mitigate the significant negative influence of computer anxiety on ease of use perceptions

Abstract

The direct or indirect influence of perceived ease of use (EOU) on user acceptance of computerised information systems is well documented. This has led to a number of studies examining system-dependent and system-independent factors that influence EOU perceptions. Among the system-independent factors, computer anxiety (CA) has been found to have a significant negative influence on EOU perceptions. In other words, users judge an application's ease of use on their level of CA. Since the negative relationship may jeopardise the acceptance of an application by some user categories, this study examines the conditions under which the relationship holds and what developers can do to mitigate the relationship so as to increase user acceptance. The information system examined is hypertext/hypermedia applications. The finding suggests that an application that surpasses the expectations of most user categories is likely to invalidate the relationship. Based on this study, a number of recommendations aimed at hypermedia developers are proposed.

Keywords:

• computer anxiety
• ease of use perceptions
• hypermedia

Notes

1. Overall responsibility: users' activities and assignments reflecting overall leadership or accountability for the system development project (e.g. project leader, responsible for requesting funds).

2. User_IS relationship: users' development activities reflected in user–designer communication and influence (e.g. initial evaluation, approval of work done, and being kept informed by IS staff).

3. Hands-on activity: the physical design and implementation tasks performed by users (e.g. defining screen layouts and report formats, user procedure manual and input output forms).

4. The Mann–Whitney statistical test uses a non-parametric statistic that tests for rank sum differences between two groups and is a conservative test that uses stringent criteria to identify significant relationships. Thus, if any relationship is identified, the result can be interpreted with more confidence (Siegal and Castellan 1988).

5. The two design heuristics not employed were (1) provide users with a sense of the entire domain at all times to mitigate cognitive overhead, and (2) make use of good/judicious colours to increase speed of reaction.