This preview shows page 1. Sign up to view the full content.
Unformatted text preview: irming the original unidimensional nature of the constructs obtained [Nunnally 19781. The factor loadim
ings structure was employed to determine the factor scores of each fr on
the five constructs.Al the scale items loaded highly on factors (constructs)
they represented, and weakly on other factors. The five factors accounted
for 7 3.3% of the total variation in the sample (see Table 1).
Since the factor scores are expressed as standardized scores, they create a
problemwhen interactions are expressed as products. Forexample, a
lowflow combination would yield a large positive term, making low/low
and high/hgh combinations indistinguishable. An interaction is interpretable
only if the lowflow combination represented by a smaller number relative
to the high/high combination. We have used a transformation suggested by CHOICE OF FOREIGN MARKET ENTRY MODE 15 Cooper and Nakanishi 119831 of the original factor scores to calculate the
interaction tern in order to resolve this problem. It involves determining a
Zeta squared value that is given by the following equation: j where Zu is t...
View Full Document
This document was uploaded on 03/22/2014.
- Summer '14