A shock to the system : the structural implications of enterprise system technology

Abstract

The last two decades have seen an increasing sophistication in the type of information systems employed by organizations. In particular we have seen the emergence of enterprise systems technology - advanced information technology specifically designed

to integrate the vast majority of an organization's processes and data flows. As the

characteristics of ES technology have encroached beyond individual user domains

and have become integrated throughout organizations, user acceptance issues have

also broadened beyond the individual unit of analysis. At the same time numerous

examples can be found both in the trade press and academic literature of organizations

wishing to use enterprise systems as a primary driver of widespread organizational

change and restructuring.

A fundamental premise of this study is that while it may be intuitively appealing to

consider technology as a primary catalyst for organizational change, it neglects to

acknowledge the presence of what is referred to as the &quoteduality of structure&quote (Giddens,

1993). Duality of structure proponents contend that while IT system protocols may

to a certain extent determine individual action, human agency can also determine the

extent to which the technology is incorporated into everyday operations. The failure

of past research to acknowledge the role of individual action and the influence of social

context in determining IT usage is considered to be a significant oversight (DeSanctis

& Poole, 1994).

Underpinned by the theory of structuration and its notion of duality, a theory of user

acceptance is put forward capable of clarifying the process by which users evaluate

and react to enterprise systems technology. The thesis reports on an empirical

investigation into the relationship between three representations of structure within an

organization: the characteristics of ES technology; job design; and social networks.

The capacity of ES technology to alter the structural elements of both job design and

social networks, and hence form user's attitudes and behavior towards the system,

is the fundamental theoretical premise of the thesis. As such this represents a clear

step forward in understanding the implications of ES technology for both users and

organizational structure.

Using a longitudinal embedded single case design, this study examines the user

acceptance and structural implications of introducing an ES into a large public sector

educational institution. A social network and job design perspective was adopted to

offer fresh insight into the dynamics of employee reaction to the introduction of ES

technology. Five hypotheses support the job design component of the thesis. It was argued that given the inherent design elements of ES technology, along with the specific

intent of the system's introduction, that users would both anticipate and perceive a

decrease in job characteristics following an ES implementation. Further, that the

positive relationship between job change and user acceptance would be moderated

by the amount of system usage reported by users. Users with a greater exposure to

the system were hypothesized to have a far stronger relationship between job change

and acceptance than low users. The ramifications of perceived or actual changes to

embedded resource exchange networks and subsequent employee reactions to those

changes were also considered. Essentially social networks were argued to play a

dual role in the user acceptance process, one being a conduit for the facilitation and

transfer of user attitudes towards new systems, the other acting as a catalyst for attitude

formation towards new systems.

Overall the findings only partially supported four of the eight hypotheses put forward.

While users were seen to anticipate an &quoteacross the board&quote decrease in job characteristics

at Time 1 following the introduction of an ES, perceived changes in job characteristics

at Time 2 were dependant on user hierarchy and the extent of system usage. Those

high in formal authority reported an increase in job enrichment following the system's

introduction, while those low in formal authority reported a decrease in overall job

enrichment. Usage was also seen to moderate the relationship between job change

and user acceptance. At Time 1 low users reported a positive relationship between

anticipated changes in meaningfulness and user acceptance. Conversely at Time 1 high

users reported a negative relationship between anticipated skill variety levels at Time 2

and user acceptance. Only one job characteristic reported a relationship between usage

and user acceptance. Low users reported a positive relationship between changes in

task identity and user acceptance. A post-hoc profile of the usage categories indicated

that high users were more likely to be a lower hierarchical position than low users.

The positive relationship reported by low users at Time 1 and Time 2 was explained by

both the nature of the system, as well as the type and quantity of information received

by low users. As senior members of the organization they were considered more likely

to receive information that highlighted its attributes in the context of their job roles. The

inherent design of ES technology, along with the specific intent it was being introduced,

facilitated largely management orientated objectives. Therefore it is unsurprising

that low users anticipating an increase in experienced meaningfulness following the

introduction of a system that enhanced their job role reported corresponding acceptance

levels. In contrast, the negative relationship between anticipated levels of skill variety

at Time 2 and perceived ease of use was explained by the affinity that high users were likely to have with the old system. To high users with a high degree of proficiency

associated with a redundant skill set, increased skill variety only represented a steeper

learning curve and an increased pressure to adapt to the new system.

The network component of the study also produced mixed results. Of the two networks

that were measured over time, only one supported the hypothesized increase in both

advice and resource exchange networks over time. Post-hoc analyses indicated that

two of the four groups exhibited network change consistent with the hypothesized

relationship. Anecdotal reports suggested that contextual elements such as geographical

location and managerial policy at a localized level determined the nature of the change

for the remaining two groups. The results failed to support the relationship between

network change and user acceptance. However, a weak but significant negative

relationship between the measure of network efficiency and user acceptance was

found. In simple terms users developing an increasingly redundant set of contacts

reported higher levels of user acceptance.

In sum, the thesis represents a contribution to enterprise systems, user acceptance

and social network literatures. In the first instance the research validates the call by

Orlikowski & Iacono (2001) to readily acknowledge the specific nature of the technology

under investigation. Despite the growth and saturation of enterprise system types,

comparatively little research has been undertaken to examine the user and organizational

issues surrounding their implementation. This research has demonstrated the capacity

for the inherent design elements of ES technology to have differential effects in terms

of job design for different user classifications. This and other findings represent a

step forward in understanding the structural and user acceptance implications of this

technology, while sign-pointing a number of promising future research avenues.

The job design results, and to a lesser extent the network efficiency results, demonstrate

the effect of social context on user acceptance. As such they provide further insight

regarding the potential determinants of user acceptance beyond the individual unit of

analysis. The findings also indicate an increasing need for user acceptance research to

stretch beyond the transitory, short term measures of user acceptance such as perceived

ease of use, usefulness, training and computer efficacy.

Finally the thesis contributes to a small, but growing literature examining the role of

social networks in the process of organizational change. In particular this thesis has

considered in detail, the attitudinal and behavioral consequences of artificially altering

established patterns of interaction. As such the study highlights the need to better

understand the role of networks not only in the case of facilitating change, but the

effect of network change in terms of change intervention success.