Correlations between the variables
Before conducting structural equation modeling analyses, we administered descriptive statistics and correlation analysis between the variables. Results are shown in Table 1.
Table 1. Correlations, means and standard deviations of the variables
|
SIAS
|
SPS
|
FNE
|
Social
avoidance
|
CPS
|
DERS
|
DMQ-R
|
Conformity motives
|
Coping motives
|
AUDIT
|
ARP
|
SIAS
|
1
|
|
|
|
|
|
|
|
|
|
|
SPS
|
.76**
|
1
|
|
|
|
|
|
|
|
|
|
FNE
|
.65**
|
.64**
|
1
|
|
|
|
|
|
|
|
|
Social avoidance
|
.73**
|
.63**
|
.49**
|
1
|
|
|
|
|
|
|
|
CPS
|
.56**
|
.57**
|
.61**
|
.48**
|
1
|
|
|
|
|
|
|
DERS
|
.55**
|
.46**
|
.50**
|
.47**
|
.45**
|
1
|
|
|
|
|
|
DMQ-R
|
.06
|
.14**
|
.20**
|
.06
|
.22**
|
.22**
|
1
|
|
|
|
|
Conformity motives
|
.22**
|
.22**
|
.29**
|
.16**
|
.25**
|
.27**
|
.56**
|
1
|
|
|
|
Coping motives
|
.13**
|
.15**
|
.21**
|
.11**
|
.26**
|
.32**
|
.66**
|
.27**
|
1
|
|
|
AUDIT
|
-.08*
|
.01
|
-.01
|
-.08*
|
.06
|
.07
|
.48**
|
.24**
|
.35**
|
1
|
|
ARP
|
.03
|
.11**
|
.06
|
.02
|
.15**
|
.17**
|
.39**
|
.25**
|
.35**
|
.90**
|
1
|
M
|
29.40
|
19.22
|
35.49
|
20.54
|
7.02
|
81.01
|
46.01
|
4.51
|
7.83
|
8.39
|
2.81
|
SD
|
14.66
|
13.72
|
8.32
|
10.45
|
5.08
|
20.15
|
14.21
|
2.04
|
3.82
|
5.88
|
3.70
|
Note. SIAS: Social Interaction Anxiety Scale, SPS: Social Phobia Scale; FNE: Fear of Negative Evaluation Scale; CPS: Concerns over physiological symptoms; DERS: Difficulties in Emotion Regulation Scale; DMQ-R: Drinking Motives Questionnaire-Revised; AUDIT: Alcohol Use Disorders Identification Test, ARP: Alcohol Related Problems.
* p < .05. ** p < .01.
|
SIAS and SPS had positive correlations with FNE, social avoidance, CPS and DERS. SIAS and SPS were positively correlated with conformity motives and coping motives. However, SIAS showed a negative correlation with the total score of AUDIT. SPS did not show a significant relationship with the total score of AUDIT, r = .01, p = .86. SIAS did not show significant correlation with ARP, r = .03, p = .49, while SPS had positive correlation with ARP.
FNE, social avoidance, CPS and DERS had positive correlations with conformity motives and coping motives. ARP did not show significant correlations with either FNE or social avoidance, all ps > .05. While CPS and DERS showed positive correlations with ARP.
Relationship between social anxiety and alcohol-related problems: Mediational effect of cognitive-behavioral-physiological characteristics of social anxiety and coping and conformity motives moderated by difficulties in emotion regulation
We administered structured equation modeling (SEM) to examine the moderated mediational effects of clinical characteristics of SA and coping and conformity motives in the relationship between SA and ARP, based on the biopsychosocial model of SAD and substance use disorders [8] (Buckner et al., 2013). In our research model, clinical characteristics of SA: FNE, social avoidance and CPS were hypothesized to predict ARP with the mediation of coping and conformity motives. Furthermore, we supposed that DER would moderate the relationship between cognitive-behavioral-physiological characteristics of SA and coping and conformity motives.
Mediation model analysis
Among the research variables in the study, FNE, social avoidance, and DER were made into latent variables via item parcelling [53].
Results from the mediation model analysis showed that the research model fits our data well, χ2 (234, N = 647) = 581.313, TLI = .953, CFI = .960, RMSEA = .048 (90% CI = .043 - .053). SA was positively correlated with FNE, social avoidance and CPS. FNE predicted conformity motives, while the path from FNE to coping motives was not significant. Social avoidance did not significantly predict conformity motives and coping motives, respectively. CPS positively predicted coping motives. However, it did not significantly predict conformity motives. Both conformity motives and coping motives positively predicted ARP. Meanwhile, a direct path from SA to ARP was not significant. In addition, direct paths from SA to conformity motives and coping motives were not significant, respectively. For the details, refer to Figure 1.
Considered collectively, results from mediation analysis suggest that SA does not directly predict ARP, but it indirectly predicts ARP. SA predicts coping and conformity motives via the mediation of cognitive and physiological symptoms. In turn, cognitive and physiological symptoms of SA predict ARP with the mediation of coping and conformity motives.
To examine the significance of indirect effects of SA, bootstrapping was conducted according to the suggestions of Shrout and Bolger [54]. The indirect effects were estimated from 10,000 samples that were randomly drawn from the raw data, and were tested in the 95% confidence interval. Results showed that direct effect of SA on ARP was not significant, while the indirect effect was significant. Indirect effect of CPS on ARP was significant. Also, indirect effect of FNE on ARP was significant. However, social avoidance did not show significant direct effect on both conformity motives and coping motives. It also did not show significant indirect effect on ARP. Direct, indirect and total effects are presented in Table 2.
Table 2. Direct, indirect and total effects between variables in the mediation model
Path
|
Direct effect
|
Indirect effect
(confidence Interval)
|
Total effect
|
SA → social avoidance
|
.82***
|
-
|
.82***
|
SA → CPS
|
.69***
|
-
|
.69***
|
SA → FNE
|
.77***
|
-
|
.77***
|
SA → coping motives
|
-.16
|
.33( .05-.63)
|
.17**
|
SA → conformity motives
|
.08
|
.23(-.08-.53)
|
.31***
|
social avoidance → coping motives
|
.05
|
-
|
.05
|
social avoidance → conformity motives
|
-.05
|
-
|
-.05
|
CPS → coping motives
|
.24**
|
-
|
.24**
|
CPS → conformity motives
|
.12
|
-
|
.12
|
FNE → coping motives
|
.16
|
-
|
.16
|
FNE → conformity motives
|
.24*
|
-
|
.24*
|
SA → ARP
|
-.07
|
.13( .07-.20)
|
.06
|
social avoidance → ARP
|
-
|
.01(-.11-.11)
|
.01
|
CPS → ARP
|
-
|
.11( .04-.20)
|
.11**
|
FNE → ARP
|
-
|
.11( .01-.22)
|
.11*
|
coping motives → ARP
|
.34***
|
-
|
.34***
|
conformity motives → ARP
|
.23**
|
-
|
.23**
|
Note. SA: Social anxiety; CPS: Concerns over physiological symptoms; FNE: Fear of negative evaluations; ARP: Alcohol related problems.
* p < .05. ** p < .01. *** p < .001.
|
Moderated model analysis
We tested moderated models under the suggestions of Preacher et al.[50]. We examined the moderation effect of DER on 1) the path between FNE and conformity motives and 2) the path between CPS and coping motives. Ping’s method [55] was applied to test the significance of the model fit and moderation effects.
Results showed a good model fit to the data. Model fit of moderation model of DER on the relationship between FNE and conformity motives (moderation model 1) was good, χ2 (39, N = 647) = 102.999, TLI = .977, CFI = .984, RMSEA = .050 (90% CI = .039 -.062). A direct path from the interaction variable (FNE x DER) to conformity motives was also significant, β = .06, p < .01. In addition, moderation model of DER on the relationship between CPS and coping motives (moderation model 2) fits the data well, χ2 (35, N = 647) = 266.024, TLI = .959, CFI = .967, RMSEA = .057 (90% CI= .050 - .065). Also, a direct path from the interaction variable (CPS x DER) to coping motives was significant, β = .12, p < .01.
To further examine the moderation effect of DER, we divided participants into two groups: high DER group (+1 SD above the mean) and low DER group (-1 SD below the mean). Results supported that the level of DER moderated the effect of FNE on conformity motives and the effect of CPS on coping motives, respectively. See Figure 2.
High DER group showed a larger moderation effect in the relationship between FNE and conformity motives compared to the low DER group, respectively, b = .07, p < .001, b = .03, p < .05. Also, high DER group showed a significant relationship between CPS and coping motives, b = .15, p < .001. While low DER group did not show a significant relationship between CPS and coping motives, b = .03, p > .05.
Moderated mediation model analysis
Lastly, we analyzed a moderated mediation model that included two interaction terms (FNE x DER, CPS x DER). The moderated mediation model fits the data well, χ2 (370, N = 647) = 1247.499, TLI = .913, CFI = .926, RMSEA = .061(90% CI = .057 - .064). See figure 3 for details.
We examined whether DER moderates 1) the indirect effect of FNE on ARP via conformity motives, 2) the indirect effect of CPS on ARP via coping motives when using the bootstrapping method. Results showed that the indirect path from interaction term of FNE and DER to ARP with the mediation of conformity motives was significant. Also, the indirect path from interaction term of CPS and DER to ARP with the mediation of coping motives was significant. Refer to Table 3.
Table 3. Direct, indirect and total effects of variables in the moderated mediation model
Path
|
Direct Path
|
Indirect Path
(Confidential Interval)
|
Total Effect
|
SA → DER
|
.59***
|
-
|
.59***
|
SA → Social Avoidance
|
.82***
|
-
|
.82***
|
SA → CPS
|
.68***
|
-
|
.68***
|
SA → FNE
|
.77***
|
-
|
.77***
|
SA → Coping motives
|
-.29*
|
.44(-.18-.72)
|
.15**
|
SA → Conformity motives
|
-.01
|
.30(-.01-.61)
|
.29***
|
DER * CPS → Coping motives
|
.09*
|
-
|
.09*
|
FNE * DER → Conformity motives
|
.14*
|
-
|
.14*
|
DER → Coping motives
|
.33***
|
-
|
.33***
|
DER → Conformity motives
|
.16*
|
-
|
.16*
|
Social Avoidance → Coping motives
|
.03
|
-
|
-.03
|
Social Avoidance → Conformity motives
|
-.06
|
-
|
-.06
|
CPS → Coping motives
|
.17*
|
-
|
.17*
|
CPS → Conformity motives
|
.09
|
-
|
.09
|
FNE → Coping motives
|
.13
|
-
|
.13
|
FNE → Conformity motives
|
.26*
|
-
|
.26*
|
SA → ARP
|
-.11
|
.16( .08-.25)
|
.05
|
CPS * DER → ARP
|
-
|
.03( .00-.07)
|
.03*
|
FNE * DER → ARP
|
-
|
.03( .01-.08)
|
.03*
|
DER → ARP
|
.08
|
.14( .09-.21)
|
.22**
|
Social Avoidance → ARP
|
-
|
-.00(-.11-.09)
|
-.00
|
CPS → ARP
|
-
|
.08( .01-.16)
|
.08*
|
FNE → ARP
|
-
|
.10( .00-.21)
|
.10*
|
Coping motives → ARP
|
.33***
|
-
|
.33***
|
Conformity motives → ARP
|
.22**
|
-
|
.22**
|
Note. SA: Social anxiety; DER: Difficulties in emotion regulation; CPS: Concerns over physiological symptoms; FNE: Fear of negative evaluations; ARP: Alcohol related problems.
* p < .05. ** p < .01. *** p < .001.
|