In this study, I examine howsupervisors’ use of information from EPM
systems relates to subordinates’ performance. Consistent with agency theory
predictions, positive associations surface for supervisory use of EPM
with task performance for different operationalizations of task performance:
through supervisory evaluations of a CSR’s performance, as well
as through call quality evaluations based onEPMsystem data by a separate
quality team. All else being equal, for each additional unit increase in supervisory
use of EPM, task performance increases by .36 (on a five-point
scale), and call quality increases by 2.97 points (on a 100-point scale).
Monitoring is considered a key mechanism to alleviate agency problems
that are represented in CWBs. Although I anticipated a negative
relationship between supervisory use of EPM and CWBs, the results
failed to support such an association. Based on agency theory, I expected
monitoring to be negatively related to OCBs. But the wider OCB
literature and some prior empirical findings of EPM research suggested
that supervisory use of EPM may be positively related to OCBs. Results
indicated a positive relationship between supervisory use of EPM and
OCBs, which align with this latter view. Although this result is desirable
from an organizational perspective, I address the finding in the general
discussion.
Study 2 has some limitations. First, itwas not possible to independently
evaluate the validity of the call-quality metrics, and so it was necessary to
defer to organizational assessments of their suitability. Also unavailable
was direct access to performance metrics, such as the average call time or
number of calls handled by CSRs. But both supervisors and quality assessment
teams had access to these metrics, which the quality assessment
teams explicitly incorporated into their evaluations of call quality.
Second, from a methodological standpoint, rather than the crosssectional
design employed here, a randomized field experiment design
would mitigate the endogeneity problem discussed earlier (Antonakis,
et al., 2010). Furthermore, because this study assesses supervisor perceptions
of how extensively they monitor subordinates, this indicator of
the frequency of monitoring is not objective. In addition, no objective
measure identifies the time between monitoring assessments or indicates
how this issue may relate to performance. Research indicates that effective
supervisors monitor their subordinates’ performance more frequently
than do noneffective supervisors (Komaki, 1986), and so precise measurement
of time lags between monitoring assessments would strengthen
these findings. Because Study 1 addresses these methodological limitations
by objectively measuring time lags between monitoring assessments,
it provides complementary evidence to Study 2 results.