Abstract

It is well known that throughout the 1980s and 1990s, with the growth of computer use in the workplace, older workers had greater difficulties than younger workers in learning new computer skills. Not only were older workers having more difficulties in acquiring new computer skills, but research revealed that their self-efficacy beliefs were also lower, having an adverse effect on their ability to learn and develop new skills. Also, there have been some conflicting results with regards to the mediating effects or influence that educational level has on the acquisition of skills. With a large baby-boom population in Canada, and despite some reaching retirement age, many older workers are choosing to remain in the workforce for various reasons. Since most jobs today require computer skills, it is important to investigate whether older workers are still having the same difficulties, since this may have important implications for organizations.
This study proposes five hypotheses to examine how age, computer self-efficacy, and educational level influence computer training outcome. Ninety-two participants, aged 18 to 66 (M = 37.6, SD = 12.7), took part in a two-hour computer training session. Demographic data were collected, followed by measures of pre-training computer self-efficacy and pretest of computer skills. Participants received computer skills instruction, and completed posttest and post-training computer self-efficacy measures.
Findings revealed that age negatively influenced computer training outcome, but no correlation was found between age and pre-training computer self-efficacy (and this remained consistent with the post-training self-efficacy measure). An ANCOVA indicated a significant effect of pre-training computer self-efficacy on computer training outcome. While age had a negative effect on training outcome, the absence of a relationship between age and pre-training computer self-efficacy ruled out the possibilities of mediation. Similarly, the absence of a relationship between age and educational level and between educational level and computer training outcome, ruled out the possibility of mediation. Post-training self-efficacy predicted both pretest and posttest results, but age had a significant negative weight, indicating that after accounting for post-training computer self-efficacy, older participants were expected to have lower posttest scores. Results are discussed in terms of implications, limitations, and future research.