Chinese version of the Recovery Self-Assessment scale: psychometric evidence from Rasch analysis and reliability estimates

<p><b>Background:</b> The Recovery Self-Assessment (RSA) is a popular and well-validated instrument for assessing recovery-oriented service around the world.</p> <p><b>Aims:</b> This study aims to develop a Chinese version of the RSA, which assesses the recovery orientation of hospital-based mental health services.</p> <p><b>Methods:</b> We conducted forward and backward translations of the RSA and modify the translated Chinese based on comments by content experts. We recruited 350 people with mental illnesses who regularly attend hospital and community mental health services. The participants completed the Chinese Recovery Self-Assessment Service User version (CRSA-SU) and convergent measures on hope and mental well-being.</p> <p><b>Results:</b> The Rasch analysis supported five of the six factors in the instrument and suggests that the “Life Goal” factor could be further split into two factors. We identified three misfit items (items 6, 12 and 17) that could be considered for removal. Both the internal consistency and test–retest reliability are between satisfactory and very good within each subscale, with the exception of the Choice subscale. The seven subscales had low positive correlations with measures of hope and mental well-being, which supported the convergent validity of CRSA-SU.</p> <p><b>Conclusions:</b> The results supported the factor structure, reliability and validity of the CRSA-SU.</p>