CFA of RSFST_METHODS

CFA of the R-SFST for patients with schizophrenia living in the community

2. Method

2.1 Participants

  Individuals with schizophrenia were recruited from three community psychiatric rehabilitation centers and Psychiatric Day Care Center at Taipei City Hospital in Taiwan in November 2014 to July 2015. Inclusion criteria were as follows: (1) diagnosis of schizophrenia according to the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-V); (2) ages more than 18 years old; (3) at least graduating from primary school; and (4) ability to understand and provide informed consent. Exclusion criteria were as follows: (1) severe brain injuries and impaired cognitive ability and (2) with diagnoses of substance use. This study was approved by the Institutional Review Board of Taipei City Hospital.

2.2 Procedures

Participants who met the inclusion criteria and signed the informed consent were assessed by the raters. The self-report R-SFST was administered in a quiet environment to groups of 6-8 participants at a time. Demographic and clinical information of the participants were collected from clinical records.

2.3 Instrument

Social Function Scale-Taiwan short version (SFST) is commonly used to measure social function in patients with schizophrenia (Song, 2001). SFST consists of 36 items divided in to seven domains: social engagement/withdraw, sociality, interpersonal, independence-competence, independence-performance, recreation, and employment subscales. However, the disadvantages of SFST were as follows: (1) high missing rates on items in SFST, (2) two subscales (independence-competence and independence-performance subscales) having ceiling effects could not discriminate patients with high social function, (3) two subscales had low internal consistency (Chiu et al., 2012). The above results would threaten the validity and reliability of SFST and limit its utility in both clinical and research settings.

  As a result, our study revised SFST into the Revised Social Functioning Scale –Taiwan short version (R-SFST). It consists of 27 items divided into six domains: 4-item interpersonal interaction, 5-item sociality, 5-item communications, 4-item leisure activities, 4-item ability of independence, and 5-item performance of independence.

2.4 Data analysis

  The statistical procedures were computed using SPSS and EQS. The factorial structure of the general linear model (GLM) was assessed through Confirmatory Factor Analysis (CFA). CFA was conducted to evaluate the 10 one-factor structure. The diagonally weighted least squares method was used to estimate CFA parameters (Forero, Maydeu-Olivares, & Gallardo-Pujol, 2009; Minrdila, 2010). We examined goodness-of-fit indices to determine the unidimensional construct of items. Four goodness-of-fit indices were used to examine the level of fit between the overall model and data, such as the ratio of chi-square value to the degrees of freedom, comparative fit index (CFI), Tucker-Lewis index (TLI), and the root mean square error of approximation (RMSEA). The criteria of a good model fit were χ2 / df<3.0, CFI>0.95, TLI>0.95, and RMSEA<0.08 (Bolle, 1989; Hu and Bentler, 1999; Bentler and Bonnett, 1980; Browne & Cudeck, 1993).

  After a domain presented a sufficient model fit, we estimated the factor loadings of the items to represent the correlation between the item and its corresponding factor. If the factor loading was <0.50, we deleted the item.

References

宋麗玉（民 90）。精神病患社會功能量表之發展與驗證-以實務應用為向。 中華心理衛生學刊，14，33-65。

Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological bulletin, 88(3), 588.

Bollen, K. A. Structural Equations With Latent Variables. 1989 New York. NY Wiley.

Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. Sage focus editions, 154, 136-162.

Chiu, E. C., Lee, Y., Lai, K. Y., Kuo, C. J., Lee, S. C., & Hsieh, C. L. (2015). Construct validity of the Chinese version of the Activities of Daily Living Rating Scale III in patients with schizophrenia. PloS one, 10(6), e0130702.

Forero, C. G., Maydeu-Olivares, A., & Gallardo-Pujol, D. (2009). Factor analysis with ordinal indicators: A Monte Carlo study comparing DWLS and ULS estimation. Structural Equation Modeling, 16(4), 625-641.

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural equation modeling: a multidisciplinary journal, 6(1), 1-55.

Minrdila, D. (2010). Mazimyum likelihood (ML) and diagnoally weighted least squares (DWLS) estimation procedures: A comparison of estimation bias with ordinal and multivariate non-normal data. Int J Digit Soc, 1, 60-66.

煩請主任及芳瑜協助editting~謝謝!!