Perceived Moderating Ability of Relational Interaction vs. Reciprocal Investments in Pursuing Exploitation vs. Exploration in RFID Supply Chains - Enterprise Information Systems Design, Implementation and Management: Organizational Applications - page 90

are organizations that had not yet implemented
RFID but are knowledgeable about it or may be
implementing RFID in the future. The specific
items used for IT infrastructure integration (i.e.,
data consistency and cross-functional application
integration) and supply chain process integration
were borrowed from Rai, Patnayakuni, and Seth
(2006); items for exploitation and exploration were
borrowed from Subramani (2004); and items for
reciprocal investments were drawn from Son, et
al. (2005) and the items for relational interaction,
from Patnayakuni, et al. (2006).
Since the organizations have not yet imple-
mented RFID, the survey respondent was asked to
indicate their perceptions of the importance of the
independent, dependent, and moderator variables
using multiple items per construct. Seven-point
Likert scales were used with minimum-maximum
anchoring points appropriate to the construct being
measured. The computer program SPSS version 15
was used in conducting a series of simple regres-
sion data analyses and their associated moderated
regression analysis runs.

To establish convergent and divergent validity,
the item-to-total correlations of the constructs were
examined and, in general, the specific items have
a stronger correlation with the construct than with
other items (Rai, Patnayakuni, & Seth, 2006).

sample profile description

The convenience sample consists of a total of
87 firms from the membership of the Council of
Supply Chain Management Professionals that
responded to a certain part of the survey ques-
tionnaire --- these were the firms that constitute
the convenience sample of organizations that are
knowledgeable about RFID or may be implement-
ing RFID in the future. About 51.06 percent of
the firms had 1,000 or less employees and 32.62
percent had more than 1,000 employees. The
following profile shows the membership of the
firms in different industry sectors: service (78.57
percent), manufacturing (21.43 percent).

Moderated regression procedure

data Measurement properties Moderated regression analysis tests whether

the relationship between two variables changes
depending on the value of another variable (i.e.,
interaction effect) (Aguinis, 2004). The mod-
erator variable explains changes in the nature of
independent variable to the dependent variable
effect, and provides information concerning the
conditions under which an effect or relationship
is likely to be stronger.
Regression analysis was conducted to test the
hypotheses presented in this study. The moderated
regression procedure requires testing first order
effects, which in this study, will be referred to as
“model 1.” A model 1 simple regression tests the
direct effects of a predictor variable on a dependent
variable. Simple regressions, therefore, were ran
between each of the independent and the dependent
variables. The variance in the dependent variable
on account of the independent variable is noted
using the R 2 value. Then, the regression procedure

The internal consistency of the items constituting
each construct was assessed using Cronbach's
alpha and the results are in conformance with
Nunnally's (1978) guidelines of getting values of
.70 or above. Generally speaking, the items have
internal consistency with values beyond the .70
threshold recommended. The different variables
used in the study showed the following reliability
results: data consistency (Cronbach alpha=.944);
cross-functional application integration (Cronbach
alpha=.930);
Financial flow integration (Cronbach al-
pha=.889); physical flow integration (Cron-
bach alpha=.942); information flow integration
(Cronbach alpha=.946); exploitation (Cronbach
alpha= .918); exploration (Cronbach alpha=.936);
reciprocal investments (Cronbach alpha=.964);
and relational interaction (Cronbach alpha=.962).