RESULTS
Each dependent measure was submitted to an ANOVA using the type of the scenario as a between-subjects
factor. The ANOVA results were not significant. Therefore, the data derived from the two scenario conditions were
pooled in the subsequent data analyses.
Hedonic value of impulse buying
To test the hedonic value hypothesis that predicts greater buying impulsiveness is associated with greater
pleasure experienced, hierarchical multiple regression analyses were performed. Pleasure rating was regressed on
impulse buying tendency and buying impulsiveness measure in a hierarchical fashion. At Step 1, the control
variable impulse buying tendency was entered. At Step 2, buying impulsiveness measure was added to the model.
A significant increment in
R
2
would indicate that the reported buying impulsiveness in the particular
purchase situation contributed to the felt pleasure after the trait-based variation was accounted for. This was exactly
the pattern emerged from the regression analysis. A significant increment in
R
2
(
∆
R
2
=
0.10
, F
(2, 275) = 33.11,
p<
0.01) was observed in the results. Regression analyses revealed that higher level of buying impulsiveness was
significantly related to higher level of felt pleasure. As predicted, the direction of the association was positive and
the magnitude of the association between the reported buying impulsiveness and felt pleasure was moderately strong
(
)
01
.
0
,
90
.
5
)
275
,
1
(
,
54
.
0
<
=
=
p
t
β
. Taken together, Hypothesis 1 was supported.
Affective ambivalence
To test the hypothesis that cognitive ambivalence experienced by an impulse buyer is positively related to
the felt affective ambivalence, hierarchical multiple regression analyses were performed. At Step 1, the control
variable impulse buying tendency was entered. At Step 2, cognitive ambivalence measure was added to the model.
A significant increment in
R
2
would indicate that cognitive ambivalence contributed to the felt affective
ambivalence after the trait-based variation was accounted for. This was exactly the pattern emerged from the
regression analysis. A significant increment in
R
2
(
∆
R
2
=
0.16
, F
(2, 265) = 26.36,
p<
0.01) was observed in the
results. Regression analyses revealed that higher level of affective ambivalence was significantly related to higher
level of cognitive ambivalence. As predicted, the direction of the association was positive and the magnitude of the
association between cognitive ambivalence and affective ambivalence was moderately strong
(
)
01
.
0
,
20
.
7
)
265
,
1
(
,
41
.
0
<
=
=
p
t
β
. Therefore, Hypothesis 2 was supported.
Dominant reactions
It was hypothesized that the outcome of cognitive evaluation would determine the dominant reactions of
the affective ambivalence experienced by the impulse buyer such that a favorable outcome of cognitive evaluation is
likely to induce a dominant reaction of pleasure while an unfavorable outcome is likely to induce a dominant
reaction of guilt.
Median split was applied on respondents’ cognitive evaluations of the impulse purchase to divide the
sample into favorable (
n
= 130) and unfavorable (
n
= 139) subgroups. Such a split was based on theoretical
Do'stlaringiz bilan baham: |