participation and information quality—user’s experiences
in current job and modeling cognition in the cognitive
communication competence construct. The former is
significant at 0.05 level, while the latter at 0.1 level.
Therefore, H2b is supported, and H3b is partially
supported. However, the beta coefficient of the interaction
term between user participation and user’s experience in
current job turned out to be negative, suggesting that the
higher a user rated their job experiences, the lower he/she
rated the information quality. A more detailed explanation
will be provided in the Discussion section. The statistics
suggest that H2a, H3a, H3c, H3d, H3e, and H4 are not
supported.
DISCUSSION In addition to the previous studies which focused
on the impact of user participation on system quality, this
study empirically suggested that user participation
significantly contributes to information quality. The
significant correlation between user participation and
information quality in hypothesis 1 implied that when
users participated in system development, the information
quality of the systems could be improved. In other words,
systems would deliver outputs that were reliable,
accurate, complete, and valuable to the users in a timely
manner. Moreover, the content of the information
delivered by the systems would be more precise, meet
users’ needs, and provide sufficient information.
This study did not find the moderating effect of
user’s experiences in using current systems or the systems
of the same type and nature on information quality.
However, the regression analysis revealed that such
experiences was a rather weak predictor to information
quality (β coefficient = 0.948, significant at 0.1 level, p-
value = 0.092). This result suggests that user’s
experiences in using current systems can partially explain
the quality of information delivered by the systems. One
explanation is when users have experiences in the current
systems or systems of the same nature, they know what
kinds of data they want to collect from the systems as
well as the informational outputs they want to see.
Therefore, they tend to be able to share with system
developers their expectations on the system requirements
and specifications.
While the regression analyses showed that users’
experiences on current systems have a positive direct
effect on information quality of a system, users’
experiences on current jobs did not. Surprisingly, it added
a negative moderating effect on the relationship between
user participation and the information quality (β
coefficient = -0.081, significant at 0.05 level, p-value =
0.05). One reason could be that because they have a lot of
experience in their current job, they tend to build higher
expectations on what new systems should deliver.
Alternatively, they resist changes made to their current
working systems. Such resistance may cause them to
perceive the quality of information delivered by the new
systems to be lower than their expectations.
Of all five types of communication competence
cognitions, the modeling cognition is the only variable
that was a significant moderator to the relationship
between user participation and information quality.
According to this result, it is reasonable to assume that a
user’s communication competence in studying the
situation and their conversational partners when first
COMMUNICATION COMPETENCE FACTORS AS MODERATORS TO THE RELATIONSHIP BETWEEN USER