Biomedical Engineering Reference
In-Depth Information
between the “new” node and the previously activated node; so the likelihood that
the new node will receive the required threshold level of activation goes up when the
spreading activation arrives via a stronger link.
Focusing only on one product category (in contrast to the across category setting
in Erdem 1998 and Erdem and Sun 2002 ), Balachander and Ghose ( 2003 ) (BG) find
empirical support for the memory-based explanation. They argue first that the parent
product can be expected to be associated strongly with the brand name (node) in
consumers' memory; consequently, when exposed to say the advertising for the
child product, the activation of the umbrella brand will likely spread to the parent
product (because of the strong link between the two). The ensuing retrieval of the
parent will lead to a positive, reciprocal spillover effect. By contrast, exposure to
advertising for the parent product is not likely to trigger the child product since the
strength of association between the child and the umbrella brand is not significant.
BG find that the reciprocal spillover effect of advertising is stronger than the parent
product's own advertising effect, and the difference in the magnitudes of these
effects is the highest when the child product has been introduced relatively recently
into the market (and goes down with time).
23.2.3
Spillover Effects in Other Contexts
The theory underlying BG's hypotheses falls under the broader Feldman and Lynch's
( 1988 ) accessibility-diagnosticity framework; BG's focus is mainly on the accessibility
(i.e., how different nodes and the associated information may be triggered by spreading
activation) part of the framework. More recent research on spillovers considers the
diagnosticity dimension of the accessed information as well.
For instance, Roehm and Tybout ( 2006 ) consider the spillover effects of a scandal
within a given product category: either the competitors of the scandalized brand
might be deemed guilty by association, or the scandal might be interpreted as unique
to the scandalized brand, thus possibly benefiting its competitors. In other words,
assuming that the relevant information for the various brands can be accessed, if the
information about the scandalized brand is perceived as being informative about
(and diagnostic for) the competing brand, then the former prediction would arise.
In an analogous spirit, Lei et al. ( 2008 ) also note that the strength of the linkage
between two nodes will depend on the direction of the activation.
Also drawing on the accessibility-diagnosticity framework, Janakiraman et al.
( 2009 ) examine whether spillover effects can arise between direct competitors in a
given product category (cf. Kumar 2005 ). They argue that since competitors may enter
the market sequentially, consumers' perceptions of the earlier entrant(s) can spillover
onto the later entrant(s). Using marketing communications and prescription choice
data from a panel of physicians, they find support for two types of spillovers: the first
is a prior-perception spillover, in which the physician draws upon the quality percep-
tions from an existing brand to form the prior perception of quality for a new entrant
(before any patients experience the new brand); the second is a dynamic-perception
Search WWH ::




Custom Search