Chinese Version of Compassion Scale Proves Reliable for Nurses

Researchers in China have validated a Chinese-language version of the Sussex-Oxford Compassion Scales (SOCS) for use with nurses, reporting that the translated instrument shows good reliability and validity in a sample of 429 nurses recruited by convenience sampling from two general hospitals in Zunyi City, Guizhou Province. The study, published in BMC Nursing, addresses a gap the authors identify directly: according to the paper, there is currently no specialized tool in China to evaluate nurses’ compassion.

The SOCS consists of two subscales, one measuring compassion for others (SOCS-O) and one measuring self-compassion (SOCS-S). With the consent of the original author, the research team translated the scale into Chinese following the Brislin model, then subjected the Chinese version to item analysis, reliability testing and validity testing using general demographic data and the translated scale.

The psychometric results were detailed in the published report. Exploratory factor analysis identified five factors within each of the two subscales, with the extracted factors explaining 73.396% of the total variance for the SOCS-S and 67.692% for the SOCS-O. Confirmatory factor analysis indicated good model fit for both subscales: for the SOCS-S, the fit indices were χ2/df = 1.452, RMSEA = 0.046, SRMR = 0.031, CFI = 0.969, TLI = 0.964 and IFI = 0.970, while for the SOCS-O they were χ2/df = 1.335, RMSEA = 0.040, SRMR = 0.035, CFI = 0.969, TLI = 0.963 and IFI = 0.969. These values fall within ranges conventionally interpreted as indicating acceptable to good fit between the observed data and the proposed factor structure.

Convergent validity was also confirmed in the study. For the SOCS-O, composite reliability (CR) ranged from 0.796 to 0.867 and average variance extracted (AVE) ranged from 0.494 to 0.620; for the SOCS-S, CR ranged from 0.847 to 0.888 and AVE ranged from 0.581 to 0.666. Content validity, assessed at the item and scale level, was likewise strong: the item-level content validity index (I-CVI) values for both subscales fell between 0.83 and 1.00, and the scale-level average content validity index (S-CVI/Ave) was 0.94 for the SOCS-O and 0.97 for the SOCS-S.

Reliability estimates reported in the paper were similarly consistent across the two subscales and their five dimensions. The Cronbach’s α for the SOCS-S total was 0.934, with its five dimensions yielding values of 0.892, 0.884, 0.870, 0.876 and 0.848. The SOCS-O total showed a Cronbach’s α of 0.895, with dimension values of 0.838, 0.847, 0.791, 0.838 and 0.839. Test-retest reliability, which gauges stability of responses over time, reached 0.908 for the SOCS-S total and 0.878, 0.824, 0.721, 0.865 and 0.714 for its five dimensions, while the SOCS-O total and its dimensions produced values of 0.892, 0.748, 0.898, 0.808, 0.719 and 0.833 respectively.

To examine convergent validity against an external measure, the researchers used Pearson correlation analysis with the Self-Compassion Scale (SCS). The correlation coefficients between the SCS and the SOCS-S, the SOCS-O and the total SOCS score were 0.530, 0.509 and 0.562 respectively, indicating moderate positive associations consistent with the scales measuring related constructs.

The authors conclude that the Chinese version of the SOCS has good reliability and validity and is suitable for measuring the level of compassion of nurses in China. They note that the instrument can help nursing administrators systematically evaluate and understand the current status of nurses’ self-compassion and compassion for others, and can provide an evidence base and theoretical foundation for developing targeted comprehensive interventions aimed at improving nurses’ compassionate ability.

The study’s scope is worth keeping in view when interpreting these findings. The sample consisted of 429 nurses from two general hospitals in a single Chinese city, recruited through convenience sampling, so the results reflect psychometric performance in that population rather than a nationally representative one. As a validation study, the work establishes measurement properties of the translated scale; it does not itself test whether any intervention improves compassion or patient outcomes. The background section of the paper frames the clinical rationale, stating that compassion helps nurses develop strong trusting relationships with patients, enables effective assessment of patients’ needs and ultimately improves clinical outcomes for patients, but those broader benefits are presented as motivation rather than as findings demonstrated by this validation work.

With those qualifications, the report offers Chinese-speaking nursing researchers and administrators a psychometrically evaluated instrument covering both compassion for others and self-compassion, filling a stated measurement gap in the Chinese nursing context. The article was published open access in BMC Nursing on 24 October 2026 by Weiyan Tian, Ruirui Pei, Xianjuan Gou, Linfei Li, Mei Chen and Hu Jiang.

Subject of Research: Validation of the Chinese version of the Sussex-Oxford Compassion Scales for measuring nurses' compassion

Article Title: Reliability and validity evaluation of the Chinese version of the Sussex-Oxford Compassion Scales for nurses

Article References: Tian, W., Pei, R., Gou, X., Li, L., Chen, M., & Jiang, H. (2026). Reliability and validity evaluation of the Chinese version of the Sussex-Oxford Compassion Scales for nurses. BMC Nursing. https://doi.org/10.1186/s12912-026-05466-2

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05466-2

Keywords: nursing, compassion, psychometrics, scale validation, self-compassion, reliability, validity, China, BMC Nursing, questionnaire, nurse wellbeing, factor analysis