Premium
Psychometric properties of the revised M alay version M edical O utcome S tudy S ocial S upport S urvey using confirmatory factor analysis among postpartum mothers
Author(s) -
Norhayati Mohd Noor,
Aniza Abd Aziz,
Nik Hazliik Hussain,
Azman Mohd Yacob
Publication year - 2015
Publication title -
asia‐pacific psychiatry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.654
H-Index - 21
eISSN - 1758-5872
pISSN - 1758-5864
DOI - 10.1111/appy.12184
Subject(s) - cronbach's alpha , psychology , confirmatory factor analysis , malay , construct validity , social support , reliability (semiconductor) , statistics , social psychology , structural equation modeling , developmental psychology , mathematics , psychometrics , power (physics) , physics , quantum mechanics , linguistics , philosophy
Abstract Introduction Social support is an essential component for the physical and emotional well‐being of postpartum mothers. The objective of this study is to determine the psychometric properties of the revised M alay version M edical O utcome S tudy ( MOS ) Social S upport S urvey using a confirmatory validity approach. Methods A cross‐sectional study was conducted involving 144 postpartum mothers attending O bstetric and G ynecology C linic, U niversiti S ains M alaysia H ospital. Construct validity and internal consistency assessment was performed after the translation, content validity and face validity process. The data were analyzed using SPSS 20.0 ( SPSS I nc., C hicago, IL , USA ) and AMOS 20.0 ( SPSS I nc., C hicago, IL , USA ). Results The original questionnaire consists of four domains (emotional/informational support, tangible support, affectionate support and positive social interaction) and 19 items. Affectionate support domain with three items only was treated as a separate construct and was not included in the factor analysis. The final confirmatory model with three constructs and 13 items demonstrated acceptable factor loadings, domain to domain correlation and best fit; (χ2[df] = 1.665 [61]; P ‐value = 0.001; T ucker– L ewis I ndex = 0.944; comparative fit index = 0.956; root mean square error of approximation = 0.068). C omposite reliability, average variance extracted and C ronbach's α of the domains ranged from 0.649 to 0.903; 0.390 to 0.699; 0.616 to 0.902, respectively. Conclusion The study suggested that the four‐factor model with 16 items (including one separate factor of affectionate) of the revised M alay version MOS Social S upport S urvey was acceptable to be used to measure social support after childbirth because it is valid, reliable and simple.