Microsoft Word 2007 ichrie conference Proceedings Final-Final 06-06-07. doc

Table 2. 
Reliability, content and convergent validity measures. 
 
Factor Loadings 
Eigenvalue 
% of Variance Cronbach 
α
 
Cognitive Dimension (CD) 
2.91 
10.40 
0.78 
CD1 
0.68
CD2 
0.74
CD3 
0.73
CD4 
0.82
CD5 
0.66
Self- Efficacy & Outcome 
expectations (SE) 
1.15 
4.10 
0.69 
SE1 
0.60
SE2 
0.72
SE3 
0.55
Relational Dimension (RD) 
1.78 
6.38 
0.62 
RD1 
0.62
RD2 
0.82
RD3 
0.68
RD4 
0.43
Ownership towards Knowledge 
(OW) 
1.19 
4.27 
0.11 
OW1 
0.49
OW2 
0.42
Propensity to Share Knowledge 
(PSK) 
1.52 
5.43 
0.85 
PSK1 
0.68
PSK2 
0.67
PSK3 
0.65
PSK4 
0.69
PSK5 
0.75
Intention to participate and share 
(BI)
8.26 
29.50 
0.83 
BI1 
0.78
BI2 
0.61
BI3 
0.74
BI4 
0.74

---

 
2007 Annual International CHRIE Conference & Exposition 
351
BI5 
0.71
Note: Items with factor loadings below 0.4 are shown here due to space limitation. OW construct was dropped due 
to very low reliability. 
 
MODEL DEVELOPMENT & HYPOTHESIS 
Based on the factor analysis, the research model was developed (see Figure 1). It is important to note that 
the dependent variable used in the study was behavioral intention (not actual behavior) of the members to participate 
and share knowledge in a described online community. As mentioned in the Section 3.1, the system was not yet 
deployed to the members prior to the commencement of the survey. It was important to learn about future users’ 
intention before the actual deployment of the system since there was evidence to indicate a strong causal link shown 
between intention and actual behavior in IS research (Venkatesh, Morris, Davis & Davis, 2003). Further, in studies 
that employ survey-based data, it is more useful to measure intention since it measures contemporaneously with 
beliefs (Agarwal & Prasad, 1999). Thus, measuring intention was considered adequate for this study. Based on 
extant literature, it was hypothesized that each of the developed constructs has an influence on member’s intention 
to participate and share knowledge in the online community. More specifically, the following hypotheses were 
tested through maximum likelihood structural equation modeling: 
H1: Cognitive dimension has a significantly positive impact on the intention to participate and share 
knowledge in the online community, 
H2: Relational dimension has a significantly positive impact on the intention to participate and share 
knowledge in the online community, 
H3: Propensity to share knowledge has a significantly positive impact on the intention to participate and 
share knowledge in the online community, and finally 
H4: Individual characteristics of low self-efficacy and high outcome expectations have a significantly 
negative impact on the intention to participate and share knowledge in the online community. 
 
Figure 1. 
The proposed research model and hypotheses
. 
Cognitive Dimension
(5 measures)
Relational Dimension
(4 measures)
Propensity
to share
(5 measures)
Low Self-Efficacy &
High Outcome
Expectations
(3 measures)
Intention to