Page 97 - Psychological Management of Individual Performance

76 work design and individual work performance
selected field experiments (e.g., Griffin, 1991; Wall, Kemp, Jackson, & Clegg, 1986)
using pre-test and post-test measurement, control and experimental groups, or both.
Kelly found support for a link between job perception, intrinsic motivation, and work
performance in only three out of the nine studies that contained data on all three vari-
ables. Productivity increases were larger with improvements in intrinsic motivation than
without, but because the number of studies in each category was small, the difference
was not statistically significant. Productivity gains were also higher when employees
perceived an improvement in job content than when they did not, although again the dif-
ference was not significant. There was no general link between perceptions of improved
job content and better job performance. Consistent with earlier reviews, and consistent
with literature on participative decision making (e.g., Miller & Monge, 1986), Kelly
(1992) concluded that the results for job satisfaction are more consistent with theoretical
predictions than the results for job performance. Kelly proposed a ‘twin-track’ model of
job redesign in which job satisfaction is suggested to be strongly influenced by percep-
tions of job content, whereas performance outcomes occur via other mechanisms, such
as improved efficiency that arises from changing work methods, work intensification,
changed expectancies (i.e., employees perceiving closer links between effort, perfor-
mance, and valued rewards), and enhanced motivation because of the improved goal
setting that often occurs at the same time as work redesign.
Since Kelly’s (1992) review, there have been only a handful of methodologically
rigorous studies that have examined the impact between work design and performance.
Wall, Corbett, Martin, Clegg, and Jackson (1990) carried out an early study on machine
operatorsmakingprintedcircuitboards.Thestudyinvestigatedtheperformanceeffectsof
training operators to diagnose the errors causing the downtime, and providing the workers
with the discretion to correct (and prevent) these faults. There was a reduction in overall
downtime after the intervention, supporting the idea that work redesign can enhance
performance. In a follow-up study, Jackson and Wall (1991) focused on the mechanisms
involved in the link between operator control and reduced machine downtime. One
mechanism is that downtime is reduced because operators can respond more quickly than
the specialists, who typically need to be called out. Another potential mechanism is that,
over time, operators learn to anticipate and prevent faults from happening, which in turn
leads to reduced downtime. Results supported the latter interpretation. Overall, there was
an initial decrease in downtime of 20%, followed by a larger delayed decrease of 70%.
The decreased amount of downtime remained at the same level between the intervention
and a one-year follow-up, yet no change was found in the downtime per incident. This
pattern of results suggested that operators had learnt to prevent the faults from occurring.
A later study (Wall, Jackson, & Davids, 1992) investigated the effects of introducing
a new payment system linked to system performance, which resulted in operators taking
greater responsibility for fixing machine faults. Even though there was no formal work
redesign, evidence for initial gains in system performance pointed to quick response,
with lagged gains attributable to operators actively anticipating and preventing faults
(Wall et al., 1992). Continuing on a similar theme, Leach, Jackson, and Wall (2000)
investigated the performance effects of giving the cutting and packaging operators greater
feedback about their performance. Feedback was enhanced through regular discussion of
faults with engineers and managers, and via the public display of information on system
performance. The number of call-outs to specialists was reduced, with a complemen-
tary increase in machine utilization, lasting at least 40 weeks after the intervention. The