Association Between Social Capital and Performance Among Core Members of Chinese Ngos in Aids Prevention and Control: A Cross-sectional Study

Background The role of non-governmental organizations in the eld of AIDS prevention and control is increasingly being valued in China and even the world.This research aimed to study the status of social capital and performance of core members of non-governmental organizations (NGOs), as well as the relation between the two, in the eld of HIV/AIDS prevention and control in China. Methods We used a multi-stage stratied cluster random sampling method to conduct a survey among 327 core members of 212 NGOs in the eld of HIV prevention and control in eight provinces in China from July to December 2015. The questionnaire included items on demographic characteristics of core members, a social capital scale, and an organizational performance scale. We analyzed the data using chi-squared test and logistic regression. Results From the perspective of social capital, the core personnel of NGOs with extensive social networks and high levels of social support accounted for a relatively low proportion of the total sample, at 29.1% (95 people) and 36.7% (120 people), respectively. The chi-squared test results indicated the statistical signicance of differences in factors such as educational background of the core personnel. The logistic regression analysis results indicated that trust in social capital was positively correlated with structural performance (OR = 2.21, 95% CI: 1.24–3.95). Social support, cohesion, and sense of belonging were correlated to the level of personal authority in performance evaluation (OR = 2.78, 95% CI: 1.43–5.39). The results also showed no statistically signicant correlation between social network in social capital and performance level. Conclusions social capital can promote better organizational For the core members of NGOs, improvement in the level of social capital and to playing an active role in HIV/AIDS and control. Our results can provide insights for decision-making and guiding the activities of NGOs engaged in HIV/AIDS prevention and control services.

The selection criteria of core members were as follows: the founders, heads, managers, and members who had made outstanding contributions to the organization, who were recommended by the heads of NGOs, and who had been involved in HIV/AIDS prevention and control services for at least one year.
With assistance from CDC staff, respondents were interviewed face-to-face by trained investigators from the Anhui Medical University after they expressed a verbal understanding of the purposes and procedures of the study and then signed consent forms. This study ensured the scienti c integrity and accuracy of the questionnaire and data through expert discussion, pilot experiment, on-site quality control, questionnaire double entry test, and other methods in the rst, middle, and later stages of the study.

Ethics statement
This study was approved by the Biomedical Ethics Committee of Anhui Medical University (No. 20131236).

Measurement of social capital
Based on the World Bank's social capital assessment tool, [24] simpli ed version of the social capital assessment tool, [25] and social support scale of Xiao, [26] we constructed the assessment tool for social capital and adapted it to the Chinese context.
We measured four dimensions of social capital in this study: social network, trust, social support, cohesion, and sense of belonging. We selected 19 commonly used items corresponding to the four dimensions. We asked the respondents to indicate their agreement using a ve-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree). The core members' score for social capital was rstly summed according to the scores for the corresponding items, and then the scores for each dimension were determined by con rmatory factor analysis, which were recorded as CmSC network, CmSC trust, CmSC support, and CmSC sense of belonging. According to the variance contribution rate corresponding to each common factor as the weight, we calculated the comprehensive statistics of social capital. The individual and total scores for each dimension were assigned as high level according to a score ³0 and low level according to a score <0, respectively, to dichotomize the measure into two-category variables. The Cronbach's α of the questionnaire was 0.80, indicating reliable internal consistency in this sample.

Measurementof performance evaluation
Through expert consultation and demonstration, as well as related pre-experiments, [27] we formulated a performance evaluation questionnaire that included the personal and organizational performance of core NGO members. We chose cooperation and authority as measures of individual performance, and nancial operations and business execution capabilities as measures of organizational performance. Cooperation can bring exibility to labor and capital markets, whereas authority can re ect the personal charm and performance of the core organizational members. Financial operations re ect the stability and sustainability of an organization, whereas business execution capabilities re ect its e ciency.
We used four questions to measure performance: (1) Do you think you have authority in the organization? (2) Do you take the initiative to seek cooperation with relevant organizations? (3) Do you think the organization has good business execution ability? (4) Do you think the organization has a good nancial operation? Response options were measured on a ve-point scale designed speci cally for this study, ranging from "very poor" (1 point) through "neutral" (3 points) to "very well" (5 points). Performance was dichotomized into two groups by taking the median of each respondents' performance as the cut-off point.
The Cronbach's alpha value was 0.92.

Other variables
We also collected information on the following variables: age (£30, 31-45, >45 years), sex (male, female), marital status (married/cohabiting, divorced/widowed, single), education (high school and below, college and above), personal background (medical, non-medical). Information on NGOs included the registration status with the Home Affairs Bureau (registered, unregistered) and prevalence of HIV/AIDS in the region served (low, medium, high) was collected as well.

Statistical analysis
First, we prepared the descriptive statistics on the sample. Speci cally, we used numerical values and percentages to describe the demographic characteristics, social capital, and performance of the participants. We employed a chi-squared test to examine the difference between organization types. The two types were grassroots NGOs and GONGOs .
Second, we established a binary logistic regression model and used the forced entry method to estimate the relation between social capital and the performance and demographic characteristics of the participants. Logistic regression and multiple logistic regression were used for single-factor analysis, and the odds ratios (ORs) and 95% con dence intervals (CIs) were calculated.
Finally, we used EpiData 3.1 for data entry and double-entry error detection, and IBM SPSS Statistics for Windows, Version 20.0 (Armonk, NY: IBM Corp.) for statistical analysis. We set the signi cance level at p <0.05.

Results
Socio-demographic characteristics of participants We collected and nally analyzed 327 valid questionnaires. Among the respondents, 219 (67.0%) were from grassroots NGOs and 108 (33.0%) were from GONGOs; most of the respondents were men (61.5%), married/cohabiting (56.8%), and with a bachelor's degree or above (65.1%). Only 32.7% of them had a medical background. Those from registered NGOs accounted for 32.7%, and those from high-prevalence areas accounted for 54.1%. Signi cant differences between grassroots NGOs and GONGOs (government-organized non-governmental organizations)in terms of marital status, education level,professional background, registration status, HIV/AIDS prevalence, and other variables were observed (P <0.05). (Table1)

Distribution of Social Capital and Performance of Participants
Regarding the four dimensions of social capital, 95 core members of NGOs (29.1%) had extensive social networks, 208 (63.6%) had a high level of trust, 120 (36.7%) had a high level of social support, and 191 (58.4%) had a strong sense of cohesion and belonging. Compared with GONGOs, grassroots NGOs had a lower proportion of core members with extensive social networks but a higher proportion with strong cohesion and sense of belonging.
In the performance indicators, the results showed that 76.8% of the core members of NGOs believed that they had organizational authority, 69.7% believed that they could take the initiative to seek cooperation, 57.8% believed that the organization's business implementation capacity was strong, and 60.6% believed that the organization's nancial operation was in good condition. We found statistical differences between grassroots NGOs and GONGOs in terms of cooperation seeking and business execution ability. (Table 2) Logistic regression Analysis The univariate model revealed that being over 45 years old could increase the level of authority of individuals in the organization. Core members who were older than 45 years, had a medical background, and worked in registered NGOs and GONGOs(Government-Organized Non-Governmental Organizations) were more likely to seek cooperation actively with relevant institutions. The following demographic variables were likewise positively correlated with the good business execution ability and nancial operation status of NGOs: older than 45 years old, married/cohabiting status, and from registered organizations.
Compared with the reference group(variable group to be referenced), social capital was generally positively correlated with performance. ( Table 3) The multi-factor model showed that an age of over 45 years, a high degree of trust, and a high level of social support were positively correlated with performance levels. Core members over the age of 45 years were more authoritative in the organization (OR = 4.37; 95% CI: 1.64-11.63) and more willing to seek cooperation actively (OR = 5.00; 95% CI: 2.10-12.17). Their organization also tended to have good business execution skills (OR = 2.88, 95% CI: 1.24-6.67). People with high trust (OR = 2.00; 95% CI: 1.15-3.47), high social support rate (OR = 1.88, 95% CI: 1.05-3.37), high cohesion and sense of belonging (OR = 1.92, 95% CI: 1.13-3.67) were more willing to seek cooperation actively. Similarly, people with high trust (OR = 2.98, 95% CI: 1.78-4.98), high social support rate (OR = 1.84, 95% CI: 1.09-3.13), people with high cohesion and sense of belonging (OR = 2.33, 95% CI: 1.40-3.88) are more likely to have good business execution skills. Regarding performance evaluation, nancial operation was only statistically signi cant in terms of the level of trust in the social capital dimension (OR = 3.42, 95% CI: 2.04-5.73). Meanwhile, we did not nd a correlation between social networks and performance (p >0.05). ( Table 4)

Discussion
Relationship between the characteristics of core members and organizational performance Our results indicated that the core members of NGOs had a low level of social capital, and the overall organizational performance level was not high. Previous research on China [28] has identi ed similar trends, indicating that the problems that existed before our investigation had not been effectively resolved.
We found that older core members in grassroots organizations had better job performance compared with younger members. The reason may be that the experience and wisdom accumulated with age could guide them to perform better. Younger core members may need to consult older core members for more experience. Meanwhile, those with higher education and professional medical degrees had a higher performance level, which also indicated the importance of education. A stable marital status also demonstrated a correlation with better organizational performance, indicating that harmony in one's family life has profound effects on working status, in line with traditional Chinese ethics. [29] NGO members from regions with different HIV/AIDS epidemic intensity reported different organizational performance levels, highlighting the impact of one's environment on work status.

Relationship between social capital and performance
We observed a certain correlation between the social capital and performance of the core members of NGOs: the higher the degree of trust and social support, the stronger the cohesion and sense of belonging of the core members, and the more bene cial to the organization's performance improvement.
We also noted a positive correlation between trust level and performance. Trust is one of the ve pillars of ethics in China and has an important impact on Chinese people's psychology and behavior. [29] Winning trust may be the basis for performing well in HIV/AIDS prevention and control. Moreover, a study [30] on the social capital of 188 hospital management committee members in Europe, according to the two dimensions of trust and value recognition, showed that a higher level of social capital correlates with a higher score for the quality of the hospital management system (AOR = 1.41). The concordance between their results and ours points to the similarities between Eastern and Western countries in the relation between trust level and performance, which can be used as a new discovery for further exploration.
Meanwhile, most core members reported a low level of social support, especially those in grassroots NGOs. This tendency may be related to the current national conditions in China, where the scale of NGOs in most communities and rural areas remains small. In some underdeveloped areas (such as the western part of China), NGOs lack funding support and human resources and rely on single nancing. They often fall into the predicament of impeded operation and development. [31] However, our results pointed to a positive correlation between social support and individual performance, as similarly reported in Laos by Sato et al. [32] They found a positive correlation between social support and the self-evaluation performance of primary health workers (AOR: 1.87; 95% CI: 1.27-2.76), thus demonstrating the positive role of social support in improving individual performance. As this problem may be common in Asian developing countries, the development of grassroots NGOs urgently requires and greater attention. Relevant departments may start with nancing channels and the deployment of human resources, taking into account material and non-material conditions, and point out the direction for the development of grassroots organizations.
The core members of NGOs usually have a "comfortable feeling," which can promote "democracy" and "sensitivity" within the organization, thereby bringing the organization together and promoting active participation. [33] These feelings explain their choice to work in non-pro t organizations. Their inherent sense of cohesion and belonging to the organization make them more serious about their work. Our results showed that a strong sense of cohesion and belonging could increase the authority, cooperativeness, and job execution ability of core members. A survey on social organizations from a university in China [34] revealed that when members of the organization identify with their team, they will show more supportive behaviors and have a higher sense of organizational identity, and even call for other people to participate in their organization. A Japanese study [35] likewise found a statistical correlation between sense of belonging and individual performance. However, the same study also found no statistical correlation between the cohesion of core members and the nancial performance of the organization. The reason for the difference in results may be the difference in management perspectives between professionals in China and Japan. Japanese organizations tend to be hampered by such issues as overly strict management thinking, poor treatment of members, and income instability. Consequently, the job satisfaction of members suffers, which could then hinder the effective operation of the organization.
We also found no statistical correlation between core members' social networks and performance, in contrast with previous research. [36] One possible explanation is that because the boundaries of social organizations in China are not clear, the core individuals of NGOs in the eld of HIV/AIDS prevention, despite having outstanding personal abilities, may inevitably rely on "personal connections" to obtain bene ts and resources, leading to them maintaining their existing social relations. [37] However, this kind of interpersonal relationship is rife with negative behavioral habits and psychological dependence that could hinder the institutionalization and standardization of social organizations in the eld of HIV/AIDS prevention and treatment in China. In other words, the perennial issue of reliance on personal connections poses risks and hidden dangers to the development of NGOs. Government regulatory authorities are thus advised to pay close attention to this issue.
Social and political theorists from Locke and Tocqueville to Putnam have emphasized the importance of social capital. [38] In China, social capital has also gradually attracted much-deserved attention. Therefore, the level of social capital of the core members of NGOs should be systematically improved. Stakeholders should consider leveraging interpersonal trust and social networks in improving the level of social support and sense of belonging of NGO members. Moreover, promoting the improvement of individual and organizational performance should be viewed as inseparable to this thrust.
Core members should also be encouraged to share social capital with non-core employees of the organization, and to cultivate and improve the stock of social capital (e.g., social networks). Non-core employees should establish good relationships with volunteers and even customers, and actively communicate with them, to maximize their personal capabilities as members of the organization. Such a direction would be equally bene cial to the social capital improvement of both individuals and organizations, as well as enable the promotion of the development of the eld and cause of HIV/AIDS prevention and treatment.
Our results can also be used to establish evidence-based models to guide organization in improving the prevention and control of HIV/AIDS.

Limitations
Despite implementing quality control measures before the eld survey and data analysis, our study still has some limitations. First, the cross-sectional research design used in our study could not derive the causal relations between the social capital and performance of core NGO members. Second, the eight provinces we studied had different economic and cultural backgrounds, which may affect the level of social capital and, therefore, the results. Finally, data on social capital and performance evaluation were obtained from the same person (self), which may have introduced bias owing to the lack of external evaluation and opinions.

Conclusions
Our research found a correlation between social capital and NGO performance. China may not be able to make full use of the role and contribution of social capital to improve the capabilities of organizational leaders and the e ciency of NGOs. The various dimensions of social capital shown in the research would help promote the development of NGOs, especially grassroots NGOs. Meanwhile, the characteristics of core personnel should not be ignored-stakeholders ought to examine this issue from the perspective of individuality. As a developmental policy tool, social capital deserves further consideration and practical research.
The authors want to thank the participants and all those who were involved in the surveys in the study. Also, special thanks to the Centers for Disease Control and Prevention in Yunnan, Sichuan, Hunan, Hubei, Shandong, Anhui, Jilin, and Gansu provinces for assistance throughout the project. We would like to thank Editage (www.editage.cn) for English language editing.

Ethics approval and consent to participate
The study was approved by the Ethical Committee of Anhui Medical University (AMUREC: 20131236) and the study team obtained informed consent from all participants. All methods were in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The participants were also well informed that completion of the questionnaire signi ed their informed consent.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.Anonymized participant data used in the preparation of this article will be made available on request from the lead author.
Contributors L S and M L reviewed the topic related literature, wrote the rst draft, analyzed the data and revised the manuscript.R X and Z B performed the study coordination and data collection. X Z performed study concept, data collection and study supervision. R C and D W performed the study concept and design, obtained funding and carried out study supervision and revision of the manuscript; guarantor for the study. All authors: checked, interpreted results and approved the nal version, contributing to the paper equally.

Consent for publication
Not applicable.

Competing interests
The authors have declared that no competing interests exist.