Differences between and within groups
Table 1 synthesizes the means, standard deviations and internal consistency (Cronbach ?)
of scores obtained by adolescents with non- binging and moderate binging.
Table 1. Descriptive analyses
Differences in the level of anxiety (STAY)
Statistical differences on state and trait anxiety between groups [F(1,79)Â =Â 30.82;
p = .0001] were found. No statistical differences between scales [F(1,79) = 1.26;
p = .27] were found. The interaction between groups and STAI-Y scales seems to be
insignificant [F(1,79) = .83; p = .37].
More specifically, post hoc test performed with Bonferroni adjustments confirmed that
the group with moderate binging presented higher scores in state anxiety (p  .001)
and trait anxiety (p  .001) than the group without binging. The group without binging
(p  .05) and the group with moderate binging (p  .05) demonstrated the same level
of state and trait anxiety.
Differences in the level of first order personality trait (EPQ)
Results revealed that participants (moderate binging vs. non-binging) did not differ
significantly on personality trait [F(1,79) = 2.15; p = .15], but the interaction
between groups and EPQ-R scales [F(3,237 = 13.25; p = .0001] and the interaction within
EPQ-R scales [F(3,237 = 3.85; p = .01] were significant.
Post hoc test performed with Bonferroni adjustments showed that the group with moderate
binging reported higher scores in psychoticism (p  .05) and neuroticism (p  .001)
than the group of adolescents without binge eating. However, the group of adolescents
without binge eating obtained higher scores in social desirability (p  .001) than
the group with moderate binging. No statistical differences in extraversion (p  .05)
between groups were found.
However, some differences were found comparing the personality traits separately in
two groups. The group with moderate binging obtained lower scores in social desirability
than psychoticism (p  .01), extraversion (p  .01), and neuroticism (p  .001).
Differences in the level of second order personality trait (IVE)
Regarding second order trait (IVE), results demonstrated statistical differences between
groups [F(1,79) = 17.47; p = .0001] and in the interaction among groups and these
aspects of personality [F(2,158) = 7.75; p = .001]. Even the interaction between impulsivity-venturesomeness-empathy
was significant [F(2,158 = 7.16; p = .001].
Specifically, the group with moderate binging had a higher score in impulsivity than
the group without binging (p  .001). No statistical differences between groups in
venturesomeness (p  .05) and empathy (p  .05) were found.
In the group of adolescents without binge eating, no statistical differences within
these trait were observed, while the group with moderate binging presented more impulsivity
than venturesomeness (p  .001) and empathy (p  .01).
Differences in the levels of eating behaviors (EDI-2)
Results demonstrated statistical differences between groups [F(1,79) = 70.17; p  .001]
such as interaction between groups and EDI-2 scales [F(10,790) = 4.70; p  .001].
The interaction between all EDI-2 scales was significant [F(10,790) = 6.04; p  .001].
In particular, post hoc tests performed with Bonferroni adjustments showed that the
group with moderate binging tended to strive for thinness more than the group of adolescents
without binge eating (p  .001). In addition to this, they showed a higher degree
of bulimia (p  .001), body dissatisfaction (p  .001), ineffectiveness (p  .001),
interoceptive awareness (p  .001), maturity fears (p  .01), asceticism (p  .001),
impulse regulation (p  .001) and social insecurity (p  .001). No statistical differences
between groups were found in perfectionism (p  .05) and interpersonal distrust (p  .05).
In Table 2 the simple comparisons were summarized within EDI-2 scales separately in moderate
binging and in group of adolescents without binge eating. Only significant statistical
comparisons within eating behaviors in both groups were reported. The group with moderate
binging presented higher scores in bulimia and impulse regulation, while the group
without binging reported higher scores in perfectionism and maturity fears (see Table 2).
Table 2. Differences between eating behaviors (EDI-2) in moderate and non-binging groups
Unlike the group of adolescents with moderate binge eating, the group without binging
reported lower scores in body dissatisfaction and the drive for thinness as compared
with the other group.
Correlation analysis and multiple regression analysis
Correlation analysis (Table 3) demonstrated that in both groups, state anxiety and trait anxiety are positively
related with ineffectiveness, interpersonal distrust, interceptive awareness, impulse
regulation and social insecurity.
Table 3. Correlations between personality characteristics and eating behaviors
In the group with moderate binging (Table 3), the analysis showed no significant correlations between eating behaviors (recognized
by the EDI-2 scales) and neuroticism. However psychoticism seems to be negatively
related only to asceticism, while extraversion seems to be positively related only
to impulse regulation. In this group, impulsivity did not relate to any eating behaviors,
while venturesomess seemed to be negatively related with ineffectiveness, interpersonal
distrust and asceticism. Empathy is positively related only to body dissatisfaction.
Conversely, in the group of adolescents without binge eating, psychoticism was positively
related to impulse regulation, while neuroticism seems to be negatively related to
ineffectiveness, interpersonal distrust, interceptive awareness, asceticism and social
insecurity.
Moreover, in contrast to the group of adolescents with moderate binge eating, in the
group of adolescents without binge eating, extraversion seems to be positively related
to body dissatisfaction, ineffectiveness, interpersonal distrust, interoceptive awareness,
impulse regulation and social insecurity. In this group, impulsivity is related to
impulse regulation, while venturesomeness seems to be negatively related with ineffectiveness
and interpersonal distrust. Empathy was not related to any eating behaviors (EDI-2
scales).
Multiple regression analysis determined the extent of the relationship between personality
characteristics and eating behaviors (EDI-2 scales). In the group with moderate binging,
trait anxiety seems to be relevant in ineffectiveness (t = 2.27; p = .01; ? = .83)
and interoceptive awareness (t = 2.13; p = .05; ? = .65), while interpersonal distrust
seems to be negatively influenced by venturesomeness (t = ?2.23; p = .04; ? = .49).
In the group of adolescents without binge eating, trait anxiety seems to be related
to perfectionism (t = 2.63; p = .01; ? = .60) and interoceptive awareness (t = 2.86;
p = .007; ? = .58). In this group, the neuroticism seems to be more relevant to perfectionism
(t = 3.16; p = .003; ? = .59), interpersonal distrust (t = ?2.17; p = .04; ? = .37)
and social insecurity (t = 2.56; p = .01; ? = .47), while the extraversion seems to
be negatively related only to interpersonal distrust (t = ?2.21; p = .03; ? = .34).
Moreover, impulsivity seems to be related to impulse regulation (t = 2.37; p = .02;
? = .39).