SEARCH

SEARCH BY CITATION

Keywords:

  • confirmatory factor analysis;
  • Japanese;
  • Liebowitz Social Anxiety Scale

Abstract

  1. Top of page
  2. Abstract
  3. METHOD
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

This study examined the internal consistency and structural/construct validity of the Liebowitz Social Anxiety Scale (LSAS) for community-dwelling subjects in Japan. A cross-sectional study that included 929 participants was conducted. Structural/construct validity was assessed on confirmatory factor analysis. The internal consistency reliability was good for the overall LSAS scale (α = 0.97) and for its original four factors (α = 0.92–0.89). The original four-factor model fit the observed data relatively better than alternative models. These findings indicate that the LSAS is a valid and reliable measure of anxiety symptoms for this community-dwelling population in Japan.

THE SELF-REPORT VERSION of the Liebowitz Social Anxiety Scale (LSAS) is a widely used self-reported questionnaire composed of 24 items. Only one study has examined the factor structure of LSAS in a Japanese population, and it extracted four factors using confirmatory factor analysis. The participants of that study, however, were university students, aged a mean of 20.6 years old.1

The objective of this study was to investigate the internal consistency and structural/construct validity of the LSAS on confirmatory factor analysis for community-dwelling subjects in Japan.

METHOD

  1. Top of page
  2. Abstract
  3. METHOD
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Subjects

A total of 929 healthy volunteers (334 male; 595 female), aged 57.8 ± 13.7 years old, who participated in the Iwaki Health Promotion Project 2010,2 completed questionnaires on their demographic background. This study was approved by the Ethics Committee of Hirosaki University.

Procedure

The self-report version of the LSAS in Japanese was given to all participants.3 The following instructions were read to the participants and reiterated as necessary: (i) this questionnaire assesses the way that social phobia (SP) plays a role in your life across various situations; (ii) read each situation carefully and answer two questions: the first question asks how anxious or fearful you feel in the situation, and the second question asks how often you avoid the situation. Please base your ratings on the way that the situations have affected you in the last week. In accordance with previous studies,2,4 the original four-factor model, two-factor model A (fear and avoidance), two-factor model B (social interaction and avoidance), one-factor model and four-factor model proposed by Safren et al.5 were analyzed on confirmatory factor analysis.

Statistical analysis

The internal consistency was evaluated using Cronbach's alpha. Three practical fit indices were used to evaluate the model fit: the root mean square error of approximation (RMSEA), the comparative fit index (CFI), and the relative fit index (RFI). RMSEA <0.05 indicates good fit. CFI and RFI close to 1 indicate a good fit. The data were analyzed using PASW Statistics-PC for Windows, version 18.0.0. and Amos-PC for Windows, version 17.0 (SPSS, Chicago, IL, USA).

RESULTS

  1. Top of page
  2. Abstract
  3. METHOD
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

The mean LSAS score was 42.4 ± 27.5. The average scores of the total fear (fear of performance, fear of social interaction) and total avoidance (avoidance of performance, avoidance of social interaction) were 23.9 ± 14.7 (11.1 ± 7.4 and 12.7 ± 7.8) and 18.5 ± 14.4 (8.3 ± 7.3 and 10.2 ± 7.7), respectively. The overall scale was found to be reliable (alpha = 0.97). Corrected item-total correlations for individual items ranged from −0.39 (item 13, avoidance) to 0.72 (item 15, avoidance). Alpha coefficients for the original four factors were 0.92 for social fear, 0.90 for performance fear, 0.90 for social avoidance and 0.89 for performance avoidance. Those of the four-factor model for fear ratings proposed by Safren et al. were 0.92 for social interaction, 0.79 for eating and drinking, 0.85 for public speaking and 0.76 for observation. Those of the same model for avoidance ratings were 0.90 for social interaction, 0.77 for eating and drinking, 0.78 for public speaking and 0.72 for observation.

The original four-factor model had the following fit indices for all subjects: RMSEA = 0.064, CFI = 0.872, and RFI = 0.825. The two-factor model A had following the fit indices for all subjects: RMSEA = 0.064, CFI = 0.870, and RFI = 0.824. The two-factor model B had the following fit indices for all subjects: RMSEA = 0.092, CFI = 0.737, and RFI = 0.681. The one-factor model had the following fit indices for all subjects: RMSEA = 0.092, CFI = 0.735, and RFI = 0.679. The Safren four-factor model for fear ratings had the following fit indices for all subjects: RMSEA = 0.088, CFI = 0.866, and RFI = 0.818. The Safren four-factor model for avoidance ratings had the following fit indices for all subjects: RMSEA = 0.080, CFI = 0.870, and RFI = 0.820. The factor loadings' interfactor correlations are given in Table 1.

Table 1. Factor loadings for Liebowitz Social Anxiety Scale items
ItemFour-factor model Safren model (fear)5
FearAvoidanceSocial interactionEating and drinkingPublic speakingObservation
PerformanceSocialPerformanceSocial
 10.597 0.523     0.626
 20.618 0.607    0.591 
 30.571 0.530   0.839  
 40.543 0.554   0.780  
 60.686 0.696    0.728 
 80.714 0.655     0.746
 90.675 0.620     0.706
130.460 0.385     0.471
140.749 0.692  0.733   
160.762 0.729    0.810 
170.654 0.632     0.635
200.710 0.689    0.703 
210.640 0.624  0.644   
 5 0.699 0.701 0.710   
 7 0.700 0.652 0.706   
10 0.757 0.721 0.752   
11 0.797 0.787 0.791   
12 0.808 0.775 0.802   
15 0.762 0.767   0.808 
18 0.703 0.665 0.704   
19 0.724 0.687 0.729   
22 0.599 0.512 0.595   
23 0.698 0.634 0.704   
24 0.592 0.479 0.593   
Interfactor correlationsPerformance fearSocial fearPerformance avoidanceSocial Avoidance Social interactionEating and drinkingPublic speakingObservation
Performance fear1   Social interaction1   
Social fear0.9581  Eating and drinking0.6631  
Performance avoidance0.7880.7531 Public speaking0.9320.5931 
Social Avoidance0.7640.7810.9661Observation0.9170.7090.9181
ItemSafren model (avoidance)5
Social interactionEating and drinkingPublic speakingObservation
 1   0.506
 2  0.592 
 3 0.765  
 4 0.830  
 6  0.710 
 8   0.711
 9   0.689
13   0.422
140.691   
16  0.775 
17   0.641
20  0.696 
210.629   
 50.692   
 70.664   
100.726   
110.786   
120.774   
15  0.805 
180.657   
190.697   
220.507   
230.640   
240.483   
 Social interactionEating and drinkingPublic speakingObservation
Social interaction1   
Eating and drinking0.6601  
Public speaking0.9330.6131 
Observation0.9120.6730.8681

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHOD
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

In this population-based study, interfactor correlations for the original four factors ranged from 0.75 to 0.97. Except for correlations of performance fear versus social fear, and performance avoidance versus social avoidance, interfactor correlations ranged from 0.75 to 0.79. A previous study using confirmative factor analysis showed a similar pattern of interfactor correlations among young university students in Japan.1 The interfactor correlations of the original four factors among patients with anxiety disorder in a Western country, however, ranged from 0.88 to 0.95, and all interfactor correlations were higher than 0.88.4 In Japan, taijin kyofusho (TKS), which is included in the DSM-IV and ICD-10 as a culture-bound variation of SP, has been reported since the early 1930s. The different response patterns may be due to ethnicity or the diagnosis of anxiety disorder.6

The present subjects were members of the public and did not directly represent patients with SP. A previous study found that shyness and SP have a number of similar features,7 but the mean LSAS score (87.0 ± 18.8) among Japanese patients with SP is double the present results.8 The factor structure of social anxiety between SP patients and the community might differ in LSAS measurement. Further investigation is needed to examine the factor structure of the LSAS among clinical subjects.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. METHOD
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES

The authors are grateful to Dr Yutaka Fukushima and Dr Hiroichi Tasaki for their helpful comments on our work. The authors declare that they have no competing interests.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHOD
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES
  • 1
    Okajima I, Kanai Y, Chen J, Sakano Y. Factor structure of the Japanese version of Liebowitz Social Anxiety Scale (LSAS): A confirmative factor analysis. Clin. Psychiatry 2007; 49: 829835 (in Japanese).
  • 2
    Funahashi K, Takahashi I, Danjo K, Matsuzaka M, Umeda T, Nakaji S. Smoking habits and health-related quality of life in a rural Japanese population. Qual. Life Res. 2011; 20: 199204.
  • 3
    Asakura S, Inoue S, Sasaki F et al. Reliability and validity of the Japanese version of the Liebowitz social anxiety scale. Clin. Psychiatry 2002; 44: 10771084 (in Japanese).
  • 4
    Oakman J, Van Ameringen M, Mancini C, Farvolden P. A confirmatory factor analysis of a self-report version of the Liebowitz social anxiety scale. J. Clin. Psychol. 2003; 59: 149161.
  • 5
    Safren SA, Heimberg RG, Horner KJ et al. Factor structure of social fears: The Liebowitz Social Anxiety Scale. J. Anxiety Disord. 1999; 13: 25370.
  • 6
    Stein DJ. Social anxiety disorder in the West and in the East. Ann. Clin. Psychiatry 2009; 21: 109117.
  • 7
    Turner SM, Beidel DC, Townsley RM. Social phobia: Relationship to shyness. Behav. Res. Ther. 1990; 28: 497505.
  • 8
    Asakura S, Tajima O, Koyama T. Fluvoxamine treatment of generalized social anxiety disorder in Japan: A randomized double-blind, placebo-controlled study. Int. J. Neuropsychopharmacol. 2007; 10: 263274.