A quasi-experimental study on a co-production- and school-based psychosocial well-being program in Hong Kong

Abstract

This study investigated the effects of “Healing Space,” a co-production- and school-based initiative aimed at promoting positive psychosocial health among secondary school students in Hong Kong. Designed according to the Self-Determination Theory, Healing Space sought to enhance positive mental health attitudes, psychological well-being, and social connectedness among participants. Using a two-wave quasi-experimental design, students completed baseline and post-test assessments using standardized measurements based on their program participation (service co-producer group: n = 171; service user group: n = 400; waitlisted control group: n = 348). The results demonstrated significant interaction effects between ‘time’ and ‘group’ on self-compassion and mental illness-related self-stigma, with participants in the service co-producer and service user group showing a faster rate of change compared to the control group. There were no statistically significant differences between the service co-producer and service user group, suggesting that Healing Space exerted a comparable influence on both groups of students. Additionally, the program did not yield statistically significant findings in indicators of social connectedness. Implications for research, practice, and policy are discussed.

Introduction

In recent years, the co-production approach has gained traction as an effective strategy for enhancing psychosocial well-being among youth^{1, 2, 3, 4, 5–6}. This research, using a quasi-experimental design, examined the effectiveness of “Healing Space,” a pilot psychosocial well-being program employing co-production for secondary school students in Hong Kong. By actively involving students in the co-creation of mental health spaces with their teachers and program facilitators, Healing Space aimed to foster positive mental health attitudes, enhance psychological well-being, and strengthen social connectedness within the school. Grounded in Self-Determination Theory⁷the initiative emphasized autonomy, competence, and relatedness, offering students a transformative experience that shifted their role from service recipients to active contributors. The following paragraphs introduce the definition of co-production, its theoretical relevance, the challenges associated with its implementation, and the design and delivery of Healing Space in local school settings.

Co-production and psychosocial well-being

Initially proposed in the public administration field for the design and delivery of public goods^8,9in the field of psychosocial well-being, the co-production approach involves a combination of activities between service providers and users, necessitating the active engagement of all key stakeholders in creating psychosocial or mental health-related programs or services^5,8,9. The first element of this approach lies in recognizing service users as assets rather than liabilities⁹. This approach shifts power from service providers to users^5,9acknowledges participants’ capabilities, and empowers them to maximize their potential and strengths through the co-created intervention process⁵.

Another radical element in co-production is the blurring of roles between service users and professionals^5,9. When service users are also recognized as professionals who have unique strengths, skills, and knowledge, professionals must consult users in service design and delivery—and users, in turn, actively contribute to shaping those services. As a result, the power hierarchy between professionals and service users diminishes, creating a level platform for meaningful collaboration among all stakeholders. Furthermore, professionals are referred to as facilitators rather than service providers, thereby changing the traditional top-down service delivery model⁵.

This leads to the third transformative element of co-production, which is the emphasis on reciprocity^5,9. As service providers and users share equal power in collaborative service design and delivery, both have to be actively involved in the process^5,9instead of placing service users under the spotlight as the “problem to be fixed,” co-production requires service users to actively contribute to the services designed and provided for the promotion of their community well-being. Hence, it underscores the reciprocal relationships between professionals and service users.

Evidence has been accumulating on the benefits of co-production-based mental and psychosocial health programs globally. Slay and Stephen (2013)⁵ found that mental health interventions using co-production were effective in promoting psychological health, enhancing social relationships, and reducing stigma towards mental illness among service users. Other youth interventions^{1, 2, 3–4} that emphasized children’s rights and actively involved young people in service co-planning were also found to reduce youth self-stigma related to mental illness and promote their mental health literacy. Additionally, a family- and intergenerational-based mental health program in Pakistan strengthened youths’ psychological resilience by inviting them to co-design a health promotion program that addressed their family-related hardships⁶.

Self-determination theory

The design and implementation of co-production-based psychosocial and mental health programs can be traced to Self-Determination Theory, which posits that autonomy, competence, and relatedness are fundamental elements of psychosocial well-being^5,7. In co-production, autonomy refers to the freedom and choice that service users have over their contributions in service co-planning and co-delivery^5,7. Competence refers to the ability to create an impact in their own lives and the lives of others through the active engagement of service users in co-designing and co-delivering services^{10, 11, 12–13}. Relatedness refers to the ability to form meaningful relationships with others for shared goals. The integration of Self-Determination Theory eventually fosters attitude change, self-compassion, and social connectedness through the co-production approach^{5,10, 11, 12–13}. Figure 1 illustrates the pathways from co-production to positive changes in mental health attitudes, self-compassion, and social connectedness through the intertwined psychological processes of autonomy, competence, and relatedness.

Fig. 1 [Images not available. See PDF.]

Conceptualizing theory of change from Coproduction to Psychosocial well-being.

Self-compassion can be cultivated by fulfilling the needs for autonomy, competence, and relatedness. Individuals experience autonomy when they are included in co-creating unique solutions that impact their lives. The inclusiveness in co-production empowers service users by respecting their unique skills and knowledge. By offering service users the freedom to make choices for themselves and improve their community’s well-being, they develop a sense of self-appreciation, leading to increased self-compassion^14,15. This sense of autonomy is tied to the sense of competence through mastery experiences in co-implementing targeted mental health services. Recognizing their unique insights and abilities confirms their belief that they are capable of creating social impact, leading to a sense of competence, which correlates with self-compassion^10,15. Co-production further offers a platform for relatedness by providing participants with opportunities to share their pains with others through the reciprocal exchange of experiences, cultivating self-compassion by knowing that “I am not alone in my suffering.”¹⁵ Moreover, the equal collaborative process of discussing and modifying targeted services with other stakeholders encourages interpersonal relatedness by embracing and understanding other people’s perspectives and experiences, thus fostering compassion for self and others.

Changes in mental health-related attitudes are driven by the co-production process in fulfilling autonomy, competence, and relatedness needs. Mastery experiences of co-producing mental health-related programs or services with other stakeholders on an equal platform lead to a sense of autonomy and competence in service users, making them influential members of their community and enabling them to make choices for themselves. These enabling experiences can disconfirm individuals’ hopeless perceptions about their current situation, thus inducing a positive attitude change regarding mental health-related matters. Furthermore, the process of co-producing a common goal encourages the exchange of perspectives among stakeholders¹⁰. Fulfilling collaborative relationships can offer new information to service users about the social world they live in, with community members embracing and accepting traditionally stigmatized mental illnesses^5,16. For example, programs that employ a co-production strategy have been shown to reduce participants’ levels of mental illness stigmatization by engaging them in exchanging perspectives with professionals and cultivating their competence in making positive changes for their community^5,10,16.

Social connectedness is fostered through promoting collaborative relationships between service users and providers and offering a social platform for service users to make decisions and impact their mental health with other stakeholders^5,7. The inclusive and equal community for co-production cultivates shared values for relatedness among stakeholders^{17, 18, 19–20}. The co-created social network that aims to strive for the common goal of mental well-being will facilitate service users’ perceived social support, foster a sense of community belonging, and contribute to the development of a close-knit community¹⁵.

Challenges in co-production

Nevertheless, challenges exist in the implementation of co-production in mental and psychosocial health programs. The most obvious barriers arise from the fundamental shifts in the dichotomous roles of service providers and users, which require extensive training¹⁴mindset changes, and humility from traditionally trained professionals. Co-production programs that lack the support and engagement of all stakeholders often result in limited impact on population well-being—and may even backfire due to clients’ unmet expectations and disappointment¹⁴. Even with the endorsement from service providers, different stakeholders may have varying expectations and understandings of their contributions, leading to confusion and potential conflicts^{21, 22–23}. Moreover, inequity exists in co-production programs. Users who are traditionally labeled as the vulnerable population often feel unsafe sharing their ideas due to the fear of being judged by others, leading to disengagement and inequality^21,22. While worthwhile, our team was mindful of these limitations and potential challenges during the design and implementation of Healing Space.

The Hong Kong context

In recent years, the mental health of students in Hong Kong has been at an alarming state, with nearly 20% of students reporting suicidal ideation and 16% struggling with mental health issues^{24, 25–26}. The global COVID-19 pandemic, the resumption of schools post-pandemic, a top-down examination-oriented education system, and typical developmental transition challenges during adolescence have all contributed to poor mental health among students^24,26,27. In Hong Kong, social stigma and fears of being labeled further prevent students from seeking help^{28, 29–30}.

Research conducted by the first author also showed that students lacked confidence and experienced fear when pursuing their desired futures, suggesting low levels of psychological resilience³¹. Information collected via pilot focus group interviews regarding school mental health further revealed that these young people did not prioritize their psychological well-being and had low levels of self-compassion³¹.

To promote school mental health and prevent further deterioration of students’ well-being, the Hong Kong government launched the “Mental Health@School” initiative, featuring the “Spread the Love, Care and Shine” campaign and a “Mental Health Literacy” online resource package for local schools³². To facilitate the design of customized school programs, the Education Bureau allocated a one-time grant to each local primary and secondary school to implement various mental and psychosocial health initiatives³³. However, an education culture that prioritizes academic success and a high level of mental illness stigmatization have made school mental health programs difficult to implement^30,34. Furthermore, these resources were offered in a top-down manner, which did not value students’ input in co-creating school mental and psychosocial health initiatives²². The professional-oriented strategy overlooked the need to view mental health issues from the students’ perspective, positioning them as the “problem to be fixed“²². To fill this service gap, we referenced current co-production literature^{1, 2, 3, 4, 5–6} to design Healing Space, a school-based and co-production-based psychosocial health program for Hong Kong schools.

The intervention: Healing space

Based on existing evidence and theoretical foundations on co-production^5,7,15we conceptualized Healing Space to change students’ mental health-related attitudes, enhance self-compassion, and promote social connectedness within the school. This is achieved by actively engaging students in co-producing positive mental health-related interactive art installations with their teachers and program facilitators. These installations focus on the theme of promoting awareness of positive mental health by providing a space for students and teachers to feel relaxed and happy (Fig. 2a and b).

Fig. 2 [Images not available. See PDF.]

Installations of Healing Space.

Program guideline

The co-production-based mental health workshops were delivered by registered social workers and program facilitators in school settings, with each session designed to achieve specific outcome (Table 1). Students who directly participated in these six workshops with the goal of co-creating their mental health art installation were identified as co-producers. These co-producers were randomly selected by their schools based on their classroom grade (i.e., grade 8, 9, or 10) as a unit for recruitment.

Table 1. Healing space session Plan.

Session Plan	Psychological Needs Fulfilled			Intended Outcomes	Participants
	Autonomy	Competence	Relatedness		Service Co-producers	Service Users Only
Session 1: Positive mental health and emotional awareness Intentions (1) Program Introduction (2) Discuss the notion of mental health with students and the importance of it. (3) Creating a safe space for self-expression, process-oriented art activities for personal choices. Check-in Movement games indicating current body status and emotions Warm-up Use images of different people (e.g., a bored child, an old lady waiting for the bus) and ask the students to guess how these people’s feelings were now and their stimulated response after seeing these images. Core Activities Options 1. Unconscious drawing for the creation of a self-portrait collage 2. Use of color to paint favorite childhood memories and feelings out of impression 3. Use words to write down the past week of happy and sad events on a very long piece of paper, students can walk around to check out their friends’ writing and share what they see. De-briefing Facilitators summarize key intentions of the workshops with the students	√			Psychological wellbeing: Self-compassion	√
Session 2: Understanding my own and my peers’ emotional needs Intentions Activities for mutual understanding and awareness of own preferences and others Warm-up: Pictionary: Students go into groups for competition to decipher an image representing and use soft clay to recreate the image for their peers to guess Core Activities Options 1. Emotional Histogram: Documenting the past weeks’ feelings and happenings, recreate it in graphical form, and share it with a classmate 2. Use color to paint past weeks’ feelings and create a mini exhibition for classroom sharing. 3. Use Legos to create a person relaxing space and share ideas in groups with their friends. De-briefing Facilitators summarize key intentions of the workshops with the students			√	Psychological wellbeing: Self-compassion Social Connectedness: Classmate Support	√
Session 3: Brainstorming positive mental health installation Intentions Advocate students’-initiated ideas and empower them to brainstorm with their peers’ positive mental health installation design. Core Activities • Students are divided into groups to share views on mental health. • Students visited the space at their schools that they need to re-design with arts installations. • Facilitators need to orient their students to link their design with positive mental health. • Students filled out a worksheet identifying their installation themes and materials needed.	√	√	√	Attitudes: Mental health awareness; Lower Mental health stigma Psychological wellbeing: Self-compassion Social Connectedness: Classmate and Teacher Support	√
Session 4: Co-designing and co-building prototype and resources gathering Intentions Co-design with students and teachers on their actual design, respect their decisions, and confirm action plans for realizing their ideas. Core Activities Open Studio for creating	√	√	√	Attitudes: Mental health awareness; Lower Mental health stigma Psychological wellbeing: Self-compassion Social Connectedness: Classmate and Teacher Support	√
Session 5&6: Co-creating and co-installing healing spaces Facilitating students to implement their projects with their group mates alongside their teachers and program facilitators for their Healing Spaces at their schools.	√	√	√	Attitudes: Mental health awareness; Lower Mental health stigma Psychological wellbeing: Self-compassion Social Connectedness: Classmate and Teacher Support; School Belonging	√
Final Exhibition Exhibition of Healing Spaces Interactive Installations at corresponding school premises		√	√	Attitudes: Mental health awareness; Lower Mental health stigma Social Connectedness: Classmate and Teacher Support; School Belonging	√	√

Note: Service Co-producers were students who participated in the six sessions of co-production workshops to co-design the mental health-related arts installation. Co-producers were recruited by school recommendations based on using classroom as a unit. Service users were students who did not participate in the workshops and only participated in using the interactive installations designed and produced by the service co-producers. All students from the schools who were not co-producers were service users as they could freely visit and interact with the mental health installations.

The six workshops began with a professionally led approach and gradually transitioned to a peer-led format with co-producers leading the production of their art installations. The making of mental health-related installations was co-facilitated by social workers and a child psychologist with expressive arts training. For more complex installations, a design specialist was hired to support the students in completing their work. At the final exhibition, the co-producers displayed their mental health-related installations in various locations within their schools for their schoolmates, i.e., service users. These service users, who did not join the six workshops, were invited to participate in the final exhibition to interact with and test out these mental health installations created by the students who served as co-producers. Both co-producers and service users received regular psychoeducation classes embedded in their school curriculum.

As illustrated in Table 1, each workshop session started with a check-in activity in which students were invited to express themselves through movements, sounds, or an image to represent their emotional or physical status, followed by the core activity. Please see Table 1 for additional details on the activities that facilitators can choose from based on their observation of students’ needs during the session.

Fidelity

We embraced the co-production approach throughout the program’s conceptualization, design, and delivery. During the initial conceptualization stage, we collaborated with a local non-profit organization known for its extensive knowledge of frontline youth practices to co-design the framework of Healing Space. At the design stage, we engaged school management and teachers to discuss each session, ensuring that Healing Space met the specific needs of individual schools and their students. This approach also ensured program fidelity by requiring all program facilitators to align themselves with the program and school expectations. The program guideline, including the intention of each session and the range of activities, was available for all program facilitators.

A standardized half-day training was offered by a registered expressive arts therapist and another half-day was offered by a design thinking specialist before the program launch to align facilitators’ expectations and attitudes. During the program, the Principal Investigators also regularly met with the team to discuss and monitor the progress of each session. An expressive arts therapist also conducted two supervision sessions to address implementation challenges and ensure program fidelity.

Adaptations and variations

On the other hand, apart from having an overall framework and protocol, we also allowed our program facilitators to adapt the program details to fit the specific needs of the students at each site and the available resources at each school. While the guiding framework of the program is informed by Western literature, Healing Space was designed to fit the local educational context. The co-production approach was specifically adopted to empower students with a sense of autonomy to support positive mental health, particularly within Hong Kong’s examination-oriented, unidirectional education system, where students often feel disempowered within local school settings. Thus, Healing Space is a locally developed program—grounded in international theory but tailored to the local context.

Research aims and hypotheses

This study examined the effectiveness of Healing Space in promoting positive mental health attitudes by decreasing participants’ self-stigma related to mental illness (Aim 1), increasing self-compassion (Aim 2), and enhancing social connectedness (Aim 3).

For self-stigma, students in the service co-producer group and service user group will exhibit a faster rate of decrease in self-stigma related to mental illness compared to those in the control group (H1).

For self-compassion, students in the service co-producer group and service user group will experience a faster rate of increase in self-compassion compared to those in the control group (H2).

For social connectedness, the service co-producer group and service user group will experience a faster rate of increase in perceived classmate support, perceived teacher support, and a sense of school belonging compared to those in control schools (H3).

Within the experimental group, co-producers will show faster rates of change in all variables of interest compared to service users (H4).

Method

This study employed a quasi-experimental design with a waitlist control group across two waves of data collection. Eight secondary schools in Hong Kong participated in the program from March 2023 to July 2024. This study was approved by The Hong Kong Polytechnic University Institutional Review Board (PolyU IRB) (IRB No.: HSEARS20230111001). All procedures were performed in accordance with relevant guidelines and regulations. Informed consent was obtained from both students and their parents before administering the survey. The study was registered with the Chinese Clinical Trial Registry (ChiCTR; Registration No.:ChiCTR2500101266).

Participants

A total of 3,061 students from 8 schools studying in Secondary 2 to 4 participated in Healing Space and filled out the research survey. Based on the preferred scheduling of the school management, five schools were assigned as experimental schools (n = 2046) while three were assigned as waitlist control schools (n = 1642). Students in the control schools attended Healing Space after the research ended. Experimental schools started the program in the first 6 months of the program and waitlist control schools started in the latter 6 months.

For the recruitment procedure, a multi-stage sampling was used to recruit the participants. Our team first worked with our community partner to randomly select 16 schools from their school network. A total of 16 schools were contacted and 8 schools agreed to join Healing Space, with 5 schools as experimental schools and 3 as waitlisted control schools. We invited management from the participating schools to use grade as a unit to randomly select students to be ‘service co-producers’ (n = 404) between secondary 2 (i.e., grade 7) and secondary 4 (i.e., grade 9). We selected these grades because secondary 2 marks the period when students have completed their first year of transition into secondary school and have become familiar with their school environment, allowing for meaningful contributions. Secondary 4 is the final year before students dedicate more time to preparing for their college entry exams in secondary 5. Students who were not selected belonged to the service user group (n = 1642). Figure 3 illustrates the CONSORT flow chart of the participants.

Fig. 3 [Images not available. See PDF.]

CONSORT (Consolidated Standards of Reporting Trials) flowchart of participants.

Measurements

The outcome variables included self-stigma towards mental illness, self-compassion, perceived support from teachers and classmates, and a sense of school belonging. All outcome variables were evaluated using standardized measurements. Scales used in this study were translated from English into Chinese by the researchers and back-translated to English to ensure construct and translation validity. These variables were time-varying variables collected in the pre-test and post-test. Demographic variables, collected at baseline, included gender (0 = female, 1 = male), age, ethnicity (0 = Not Chinese, 1 = Chinese), and financial satisfaction (1 = Not satisfied at all, 10 = Very satisfied).

Self-stigma related to mental illness

Self-stigma related to mental health was assessed using the self-stigma subscale of the Stigma and Self-Stigma Scales (SASS)³⁵. This subscale consists of 6 items that evaluate personal beliefs about oneself if experiencing a mental health problem rated on a 5-point Likert scale from 1 (Strongly disagree) to 5 (Strongly agree). This scale showed good reliability and construct validity with the current students’ sample (Cronbach’s α = 0.90, CFI = 1.00; TLI = 0.98).

Self-compassion

Self-compassion was measured using eight items adapted from the Self-Compassion Scale for Youth (SCS-Y)³⁶. The original scale consists of 17 items on a 5-point Likert scale ranging from 1 (Almost never) to 5 (Always). Eight items from the common humanity, self-kindness, and over-identification subscales were selected as pertaining to the program intention. We omitted some items to avoid overloading the participants with a long survey consisting of irrelevant questions. This scale demonstrated good reliability and construct validity with the current sample (Cronbach’s α = 0.82, CFI = 0.98, TLI = 0.96).

Perceived teacher and classmate support

The Teacher and Classmate Support Scale was used³⁷. This scale employs a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) and evaluates two dimensions of perceived social support: teacher support and classmate support. Each dimension comprises four items. The scale showed good reliability and construct validity using a two-factor model approach with the current sample (Teacher support Cronbach’s α = 0.90 and classmate support α = 0.89; CFI = 0.99, TLI = 0.99).

School belonging

School belonging was evaluated using the Sense of School Belonging Scale from the OECD Programme for International Student Assessment³⁸. This scale comprises six items that measure the feeling of acceptance and belonging within a community on a 4-point scale (1= ‘strongly disagree’ to 4 = ‘Strongly agree’). The scale demonstrated good reliability and construct validity with the current sample (Cronbach’s α = 0.72, CFI = 0.99, TLI = 0.97).

Data analysis

Descriptive statistics were conducted to explore the collected information. A one-way ANOVA and Chi-square test were used to explore the baseline differences. A mixed-effects repeated measures ANCOVA was employed to test the program’s effectiveness. Outcome variables were time-varying variables, which were: self-stigma towards mental illness, self-compassion, perceived teacher and classmate support, and sense of school belonging. Predictors included the within-group variable of Time (Baseline = 0 and Post-test = 1) and between-group variable of participation levels (Waitlisted Control group = 0; Service user group = 1; Service co-producer group = 2). To examine the different rates of changes in the outcome variables over time among groups, the interactions between “time” and “group” were examined. Prior studies indicate that demographic factors such as gender, ethnicity, financial satisfaction, and age may influence intervention effectiveness on mental health³⁹. Therefore, these variables were included as covariates. Listwise deletion was used to handle missing data. All statistical analyses were performed using IBM SPSS 26.0.0.

Results

Descriptive findings

A total of 919 students (service co-producer group n = 171, service user group n = 400, waitlisted control group n = 348) completed both the pretest and posttest and were included in the final analysis (mean age = 14.55; female students = 46.9%; male students = 53.1%; ethnically Chinese = 88.58%). The means of all outcome variables in the pre-and post-test for the three groups were calculated (Table 2). Results of one-way ANOVA indicated no significant differences among the groups at baseline, except for age and mental illness-related self-stigma. For missingness, 57.67% of the service co-producers, 64.70% of service users, and 38.73% of those in the waitlist control group dropped out after the first wave of data collection.

Table 2. Descriptive statistics and test of mean differences of outcomes and Covariates.

		Group 1 (providers) n = 171		Group 2 (users) n = 400		Group 3 (Control) n = 348
Variables		M(SD)ba	M(SD)Post	M(SD)ba	M(SD)Post	M(SD)ba	M(SD)Post	Fba/Χ2ba
Outcomes
self-compassion	8–40	25.11(4.88)	27.59(6.88)	25.00(5.41)	26.39(6.12)	24.70(6.90)	25.09(6.89)	0.37
Self-Stigma (Mental Health)	0–24	12.04(4.88)	9.38(6.13)	12.22(5.10)	9.2(6.72)	10.84(5.63)	9.84(5.67)	6.86*
Teacher support	4–20	14.29(3.46)	14.2(3.69)	13.87(3.40)	14.45(3.61)	14.37(4.04)	14.98(3.93)	1.96
Classmate Support	4–20	14.39(3.43)	14.76(3)	14.09(3.45)	14.12(3.7)	13.91(3.80)	14.28(3.84)	1.04
School Belonging	6–24	17.15(3.5)	17.08(3.4)	16.72(3.28)	17.11(3.19)	16.43(3.02)	16.87(3.42)	2.9
Covariates
Age	11–20	14.55(1.24)	--	14.29(1.67)	--	14.84(1.73)	--	10.88*
Finance satisfaction	1–10	6.71(2.08)	--	6.57(2.23)	--	6.55(2.42)	--	0.29
% female	--	40.24	--	48.72	--	48.13	--	Χ2ba = 3.740
% Chinese	--	86.81	--	89.76	--	89.94	--	Χ2ba = 2.044

*p < 0.05. ba, Baseline.

Repeated measures covariance analysis (ANCOVA)

Table 3 shows the results of the repeated measures ANCOVA by the outcome of interests.

Table 3

Results of repeated measures ANCOVA for individual-based variables.

Outcome Variables	Self-Stigma		Self-compassion		Teacher support		Classmate Support		Sense of School Belonging
	F		F		F		F		F
Time	F (1,831) = 1.88	< 0.01	F (1,831) = 0.50	< 0.01	F (1,834) = 0.01	< 0.01	F (1,834) = 0.05	< 0.01	F (1,833) = 2.16	< 0.01
Group	F (2,831) = 1.30	< 0.01	F (2,831) = 7.04*	0.02	F (2,834) = 2.19	0.01	F (2,834) = 1.24	< 0.01	F (2,833) = 1.91	< 0.01
Gender	F (1,831) = 1.07	< 0.01	F (1,831) = 1.27	< 0.01	F (1,834) = 0.04	< 0.01	F (1,834) = 0.24	< 0.01	F (1,833) = 0.06	< 0.01
Age	F (1,831) = 2.65	< 0.01	F (1,831) = 21.30*	0.02	F (1,834) = 7.28*	0.01	F (1,834) = 5.44*	0.01	F (1,833) = 0.93	< 0.01
Ethnic	F (1,831) = 0.01	< 0.01	F (1,831) = 0.77	< 0.01	F (1,834) = 3.42	< 0.01	F (1,834) = 5.88*	0.01	F (1,833) = 10.23*	0.01
Finance satisfaction	F (1,831) = 1.42	< 0.01	F (1,831) = 23.83*	0.03	F (1,834) = 50.93*	0.06	F (1,834) = 20.51*	0.02	F (1,833) = 21.21*	0.02
Time*Group	F (2,831) = 6.16*	0.02	F (2,831) = 4.44*	0.01	F (2,834) = 1.75	< 0.01	F (2,834) = 1.01	< 0.01	F (2,833) = 0.95	< 0.01
Time*Gender	F (1,831) = 7.74*	0.01	F (1,831) = 1.44	< 0.01	F (1,834) = 0.04	< 0.01	F (1,834) = 0.14	< 0.01	F (1,833) = 1.18	< 0.01
Time*Age	F (1,831) = 2.43	< 0.01	F (1,831) = 0.07	< 0.01	F (1,834) = 0.12	< 0.01	F (1,834) < 0.01	< 0.01	F (1,833) = 0.78	< 0.01
Time*Ethnic	F (1,831) = 3.60	< 0.01	F (1,831) = 3.27	< 0.01	F (1,834) = 16.50*	0.02	F (1,834) = 10.85*	0.01	F (1,833) = 0.01	< 0.01
Time*Finance satisfaction	F (1,831) = 4.10*	0.01	F (1,831) = 0.64	< 0.01	F (1,834) = 31.26*	0.04	F (1,834) = 21.03*	0.02	F (1,833) = 13.23*	0.02

*p < 0.05.

Self-stigma related to mental illness

The results of the mixed repeated measures ANCOVA on self-stigma demonstrated a significant interaction effect between ‘time’ and ‘group’ (F (2,831) = 6.16^*, = 0.02). The results of simple effects analysis showed that self-stigma significantly decreased over time in all groups (see Table 4). However, service co-producer group (Δ_pre−post = 3.13, F (1,808) =26.22^*, = 0.03) and service user group (Δ_pre−post = 2.70, F (1,808) =41.66^*, = 0.05) exhibited significant more reduction in self-stigma compared to the and waitlisted control group (Δ_pre−post = 0.97, F (1,808) =5.43^*, = 0.01), indicating program effectiveness. Figure 4; Table 4 shows the changes over time among participants with different engagement levels.

Table 4

Results of simple effect analysis for Self-Stigma and Self-compassion.

	Self-Stigma				Self-compassion
	Mean Difference T0−T1	F		P	Mean Difference T0−T1	F		P
Group 1 (provider)	3.13	F (1,808) = 26.22*	0.03	< 0.001	−2.76	F (1,831) = 17.31*	0.02	< 0.001
Group 2 (user)	2.70	F (1,808) = 41.66*	0.05	< 0.001	−1.55	F (1,831) = 12.01*	0.14	0.001
Group 3 (Control)	0.97	F (1,808) = 5.43*	0.01	0.020	−0.42	F (1,831) = 0.85	< 0.01	0.358

*p < 0.05.

Fig. 4 [Images not available. See PDF.]

Changes in self-stigma among participants with different engagement levels.

Self-compassion. The results showed significant main effects of Group (F (2,831) = 7.04^*, =0.02), revealing that students in the service co-producer group demonstrated higher overall levels of self-compassion. The results also showed a significant interaction effect between ‘time’ and ‘group’ (F (2,831) =4.44^*, =0.01) in self-compassion. For the waitlist control group, there was no significant change over time (Δ_pre−post=−0.42, F (1,831)=0.85, < 0.01). While there were significant changes in the service co-producer group (Δ_pre−post=−2.76, F (1,831) =17.31^*, =0.02) and service users group (Δ_pre−post=−1.55, F (1,831) =12.01^*, =0.01), indicating program effectiveness (Tables 4and Fig. 5).

Fig. 5 [Images not available. See PDF.]

Changes in self-compassion among participants with different engagement levels.

Further analysis within two experimental groups (“service co-producers” versus “service users”) on self-stigma and self-compassion indicated that there was no significant interaction effects associated with the engagement levels and time (see Appendix 1). This suggested that the effects of the intervention were consistent across all students in two experimental groups.

Perceived teacher support, perceived classmate support, and sense of school belonging

The results of the mixed repeated measures ANCOVA indicated that neither the main effects of ‘time’ or ‘group’ nor the interactions between ‘time’ and ‘group’ were significant for these variables of interest.

Covariates

Results indicated that age had a positive effect on self-compassion and perceived teacher and classmate support. Chinese students exhibited significantly higher levels of perceived classmate support and school belonging compared to non-Chinese students. On the other hand, ethnicity had statistically significant interaction effects with time on perceived teacher and classmate support, with non-Chinese students experiencing faster rates of positive change over time. Financial satisfaction had significant main effects on all variables except for self-stigma. Additionally, financial satisfaction had significant interaction effects with time on all variables except for self-compassion. Students with higher levels of financial satisfaction showed better outcomes and a faster change over time.

Discussion

This study evaluates the effectiveness of a school-based, co-production-based psychosocial health program, Healing Space, in enhancing students’ positive mental health attitudes, self-compassion, and social connectedness. The evaluation employs a two-wave quasi-experimental design, including a service co-producer group (n = 171), a service user group (n = 400), and a waitlisted control group (n = 348). Results indicate statistically significant reductions in self-stigma related to mental illness and increases in self-compassion among students in the experimental groups. No significant changes are observed in the other outcomes. Further analysis reveals no significant differences in the effectiveness of Healing Space between the service co-producer group and the service user group.

The statistically significant change in self-compassion levels among the participants can be attributed to the program’s focus on self-autonomy, mutual respect, and equality throughout all workshop sessions^5,7,9. Program instructors consistently emphasized the importance of recognizing one’s own preferences and emotional needs through direct advice and co-creating mental health installations with the students. From the first to the last session, facilitators respected students’ decisions and communicated constructively about the feasibility of their ideas^5,9. These key elements of co-production foster self-compassion through respect and recognition^5,15. Additionally, the experience of co-creating mental health installations for their peers is an act of self-kindness among the students, highlighting the importance of prioritizing their needs, which further enhances their self-compassion¹⁵.

The reduction in self-stigma related to mental illness suggests that the mastery experiences in Healing Space may have contributed to this change^7,10. Attitude transformation is often more effective through practical applications involving discussions and real-life experiences¹⁰. In Healing Space, students were invited to take the lead with their peers in co-creating installations for all key school stakeholders, including teachers, janitors, and school administrators⁵. The shift in students’ roles to service providers facilitates their understanding of mental health from the perspective of service providers, resulting in reduced self-stigma related to mental illness¹⁰. Additionally, as students become more self-compassionate throughout the workshops, they also become more accepting of their own issues, including potential mental health problems, leading to a decreased level of self-stigma related to mental illness¹⁵.

Notably, this study did not find statistically significant changes in perceived teacher support, perceived classmate support, or sense of school belonging. The non-statistically significant findings for perceived classmate support may be attributed to inadequate adult support within the program. Intensive and timely guidance from adults is essential to foster social connectedness in co-production programs, particularly when students encounter conflicts during group discussions about their mental health installations²². Additionally, post-program focus group interviews revealed that some students felt they were not receiving sufficient group-based attention and support from program facilitators in the final two sessions, leading to their disengagement and feelings of being lost in the process.

Some participants also reported not receiving adequate support from their teachers throughout the program. Their experiences align with our team’s observation of limited endorsement of Healing Space by certain teachers and school management. The absence of support from key stakeholders in the targeted services can undermine the intended outcomes of co-production-based programs^14,23contributing to the lack of improvement in perceived teacher support and a sense of school belonging.

Contrary to our hypothesis, the results reveal no significant difference in the effectiveness of this intervention between service co-producers and service users. The findings suggest a possible spillover effect of Healing Space by directly involving only some students as co-producers. Additionally, the mental health-related spaces and installations created by the co-producers might influence their peers. However, further investigation is needed to verify this assumption and map the theory of change. For example, feedback forms could be created for users to rate the mental health installations designed by their peer co-producers. School environment surveys could also be distributed to examine whether these student-designed mental health installations contribute to a more inclusive school atmosphere, thereby enhancing psychosocial well-being.

Regarding confounding effects, older students may have greater cognitive maturity, enabling them to embrace themselves and better connections with their teachers and classmates due to their longer tenure at schools. At baseline, it is understandable that Chinese students exhibited a higher sense of belonging compared to non-Chinese students. However, non-Chinese students experienced a faster rate of increase in their perceived teacher and classmate support over time, which may be related to their ability to make friends and overcome cultural barriers as they become more familiar with their schools. Additionally, students with higher levels of financial satisfaction demonstrated better outcomes and faster rates of positive changes over time. This may be attributed to social inequality, where students with higher financial satisfaction may come from more affluent families, contributing to better psychosocial well-being.

Limitations

The findings of this study should be interpreted in light of the following limitations in research and intervention design. First, we did not employ a randomized controlled trial (RCT) to examine the effectiveness of Healing Space, which is considered the gold standard for evidence-based practice. Instead, we opted for a quasi-experimental study due to the challenges of additional administrative procedures that could further deter schools from participating. Therefore, our team decided to follow the schools’ preferences and schedules in implementing the research and program.

Furthermore, compared to other art-based interventions among youth, the response rate in the post-test survey in this project is relatively high⁴⁰. As Healing Space was delivered during class periods, students would need to notify their teachers if they wished to discontinue participation, and the school had to let our team know about it. Based on our communication at schools, all students that joined the workshop were engaged in all sessions, without encountering the problem of dropout. For the service user group, their participation is defined by their interactions with the mental health installations at their campuses, hence, they are free to choose whether to interact with the installations or not at their own time and own preferences. Hence, it is reasonable to conclude that the high attrition rate was due to the low response in the post-test survey. Upon receiving feedback from the teachers at the participating schools, the low response rate was related to the lack of incentives to fill out the survey. Unlike the practices of conducting the pre-test, our team also did not allocate a designated time for the students to fill out the post-test survey in one setting. Students were given the survey in paper and pen form or via our online platform and returned it at their earliest convenience, which resulted in difficulty in filling out the survey. Nevertheless, the relatively high attrition rate could have led to biased findings associated with participants’ disengagement, which we speculate to be associated with the perceived lack of support from teachers in the program.

Moreover, the absence of a follow-up test makes it difficult to ensure the sustainability of self-compassion and self-stigma related to mental illness over time. Given the high attrition rate in the post-test, we anticipated a very low response rate if a follow-up test was implemented. In terms of the intervention, the absence of a teacher training component in Healing Space may have contributed to the disengagement of teachers and school management, which could have affected the effectiveness in improving students’ perceived social connectedness. Limited group-based support from our facilitators in the last two sessions of the program might have caused frustration among the students, leading to program ineffectiveness in the social connectedness indicators.

Last, we did not include the key psychological mediators of sense of autonomy in the questionnaire, and hence, were not able to capture the change of the students in their perceived sense of autonomy. This construct was excluded due to concerns about questionnaire length, with school management suggesting that the team remove this set of scales to avoid overburdening students. We also did not conduct a path analysis identifying the underlying mechanism from coproduction to the outcome of interests via the self-determination theory constructs. Further quantitative research is needed to verify these pathways and refine the theory of change proposed in this study.

Implications

The study has significant implications for theory, practice, and educational policy. Theoretically, this study integrates self-determination theory with co-production approaches⁵ and applies this framework to fulfill youths’ specific developmental needs for autonomy, competence, and social connectedness. Our findings delineate the mechanism of co-production strategies, suggesting a theory of change that highlights how program input enhances self-compassion and decreases mental illness-related stigma. While further research is needed to map the pathways from co-production to various aspects of psychological well-being, this quasi-experimental study facilitates knowledge transfer by demonstrating that Healing Space, a theoretically referenced co-production intervention, can impact psychosocial health promotion.

Practically, this study provides guidelines for implementing co-production-based programs in schools. The findings advocate for a bottom-up, strengths-based intervention design that empowers students to be service providers and encourages them to take charge of their own health. While further study is necessary to examine the effectiveness of each workshop to identify best practices, our study suggests the high feasibility of implementing co-production programs in school settings despite previous challenges mentioned in the literature^14,22,23.

Additional engagement is necessary to encourage active participation from teachers and school management, ensuring that students feel supported during the co-production journey. Based on our preliminary findings and observations, it is reasonable to speculate that schools with greater teacher endorsement and support and more input from the teachers can have a better effect on the students’ outcomes. In addition, schools that can afford an indoor space exclusively for the students to decorate with installations and sustain the installations have a long-term effect on the students. Follow-up visits at some of the schools with indoor permanent spaces show that these schools are able to continue to utilize the space to spin off new ideas, such as organizing a peer caring team and adding a pet corner in the “Healing Space” co-designed by the students and the teachers (Fig. 2a and b). Further quantitative and qualitative follow-up research will need to verify the observations.

Furthermore, more intensive facilitation should be provided for the students, especially during the final stage of co-production, where students need to work closely together to generate the targeted products. These insights have facilitated us in further refining this pilot intervention to promote positive mental and psychosocial health at the school level^32,33. Our findings also inform the design of a train-the-trainer program, where school management can adopt our protocol and initiate their co-production-based mental health initiatives based on our program values and theory of change.

Future peer training programs can be implemented at the school level. By engaging a smaller group of students as co-producers to offer psychosocial services to their peers, a positive school mental health culture can be fostered. For instance, school management can consider setting up a peer-led emotional health club or peer counseling club supervised by teachers. Another example is inviting students to co-create a permanent space for both teachers and students at their schools for relaxation and ‘me-time,’ emphasizing self-compassion and social connectedness. These suggestions, stemming from our current research, offer a sustainable approach to promoting positive mental and psychosocial health in school settings over the long term.

The findings also carry important implications for school mental health policy. First, to effectively promote positive school mental health, we need to offer training that helps educators shift their mindsets and recognize their students’ potential. For example, educators should not view students as problems to be fixed, but rather respect students’ preferences, needs, and strengths. Incorporating co-production into pre-service teacher training may help transform the culture of school mental health. Healing Space can be developed into a co-production-based emergent mental health curriculum for secondary schools, allowing students to actively contribute to the syllabus and workshops, fostering ongoing engagement and dialogue between students and teachers. The establishment and normalization of such a curriculum will contribute to a shift in school culture where students’ voices are respected and valued. Pilot programs based on the Healing Space model can also be adapted for tertiary, primary, and even kindergarten settings to promote the psychological well-being of children and young people across different developmental stages and educational environments.

Conclusion

This study investigates the effectiveness of Healing Space, a school-based and co-production-based program aimed at promoting positive psychosocial health in secondary schools in Hong Kong. Based on Self-Determination Theory, Healing Space was designed to improve positive mental health attitudes, psychological well-being, and social connectedness among student participants. Results indicate the program’s effectiveness in increasing self-compassion and reducing self-stigma related to mental illness, but not in enhancing social connectedness. Further findings reveal no differences in effectiveness between the two experimental groups, suggesting that the project had a comparable impact on both co-producers and service users. The findings carry important implications for research, practice, and policy in building positive school mental health.

Author contributions

A.H.Y.L. conceived and designed the study, prepared materials, drafted the manuscript, facilitated activities, and managed the project. S.J.W. analyzed the data and drafted the manuscript. K.W.Y.W. prepared materials, collected data, and analyzed data. X.Z. and J.S. collected data and facilitated activities. All authors reviewed the manuscript.

Funding

This study was funded by Hong Kong Health Bureau, Mental Health Initiatives Funding Scheme (Phase 2).

Data availability

Due to the restrictions from the University’s requirements, data cannot be shared openly but is available on reasonable request from the correspondence author.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval

This study was approved by The Hong Kong Polytechnic University Institutional Review Board (PolyU IRB) (IRB No.: HSEARS20230111001). All procedures were performed in accordance with relevant guidelines and regulations.

Consent to participate

Informed consent was obtained from both students and their parents prior to survey administration.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Bell, J et al. Nothing about Us without Us: Co-production ingredients for working alongside stakeholders to develop mental health interventions. Adv. Ment Health; 2023; [DOI: https://dx.doi.org/10.1080/18387357.2021.2020143]

2. Girma, E et al. Facility based cross-sectional study of self stigma among people with mental illness: towards patient empowerment approach. Int. J. Ment Health Syst.; 2013; [DOI: https://dx.doi.org/10.1186/1752-4458-7-21] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24004512][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3848304]

3. McMellon, C et al. Reconceptualising coproduction as activism together. Child. Soc.; 2024; [DOI: https://dx.doi.org/10.1111/chso.12788]

4. Mittal, D et al. Empirical studies of self-stigma reduction strategies: A critical review of the literature. Psychiatr Serv.; 2012; [DOI: https://dx.doi.org/10.1176/appi.ps.201100459] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22855130]

5. Slay, J. & Stephens, L. Co-production in Mental Health: A Literature Review (New Economics Foundation, 2013).

6. Vostanis, P et al. Youth-led co-production of mental health promotion in pakistan: intergenerational influences. Health Promot Int.; 2024; [DOI: https://dx.doi.org/10.1093/heapro/daae010] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38381915][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10880878]

7. Ryan, RM; Deci, EL. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. Am. Psychol.; 2000; [DOI: https://dx.doi.org/10.1037/0003-066x.55.1.68] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11392867]

8. Brandsen, T; Honingh, M. Distinguishing different types of coproduction: A conceptual analysis based on the classical definitions. Public. Adm. Rev.; 2016; [DOI: https://dx.doi.org/10.1111/puar.12483]

9. Glynos, J; Speed, E. Varieties of co-production in public services: time banks in a UK health policy context. Crit. Policy Stud.; 2012; [DOI: https://dx.doi.org/10.1080/19460171.2012.730760]

10. Bandura, A. Social Learning Theory (Englewood Cliffs, 1977).

11. Keys, C. B. et al. American Psychological Association,. The what, the how, and the who of empowerment: Reflections on an intellectual history. in APA handbook of community psychology: Theoretical foundations, core concepts, and emerging challenges (eds. Bond, M. A., Serrano-García, I., Keys, C. B. & Shinn, M.) (2017). https://doi.org/10.1037/14953-010 213–231.

12. Rappaport, J. & Seidman, E. Handbook of Community Psychology (Kluwer Academic/Plenum, 2000). https://doi.org/10.1007/978-1-4615-4193-6

13. Sen, A. Nussbaum, M; Sen, A. Capability and well-being. The Quality of Life; 1993; Oxford University Press: pp. 30-53.

14. Mayer, C; McKenzie, K. … it shows that there’s no limits’: the psychological impact of co-production for experts by experience working in youth mental health. Health Soc. Care Community; 2017; [DOI: https://dx.doi.org/10.1111/hsc.12418] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28695628]

15. Neff, KD. Self-compassion: theory, method, research, and intervention. Annu. Rev. Psychol.; 2023; [DOI: https://dx.doi.org/10.1146/annurev-psych-032420-031047] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35961039]

16. Sims-Schouten, W; Maynard, E; Pound, M. I hate having my mental health - Making sense of mental health through coproduction and visual methods with young people with complex needs. J. Youth Stud.; 2024; [DOI: https://dx.doi.org/10.1080/13676261.2022.2101358]

17. Fledderus, J; Brandsen, T; Honingh, M. Restoring trust through the co-production of public services: A theoretical elaboration. Public. Manag Rev.; 2014; [DOI: https://dx.doi.org/10.1080/14719037.2013.848920]

18. McMillan, DW; Chavis, DM. Sense of community: A definition and theory. J. Community Psychol.; 1986; [DOI: https://dx.doi.org/10.1002/1520-6629(198601)14:1<6::AID-JCOP2290140103>3.0.CO;2-I]

19. Plumb, S. J. How might the experiences and perspectives of community participants collaborating in socially engaged Art practicesbe heard, considered and equally valued alongside those of the Artist and gallery? What ethical implications for the Artist and thegallery arise when community participants’ experiences are considered in this way? PhD thesis (University of Leicester, 2017).

20. Vanleene, D., Voets, J. & Verschuere, B. To empower & include: how professionals impact co-productive community development. In IIAS Workshop on Coproduction (2019).

21. Dewa, LH et al. Reflections, impact and recommendations of a co-produced qualitative study with young people who have experience of mental health difficulties. Health Expect.; 2021; [DOI: https://dx.doi.org/10.1111/hex.13088] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32515538]

22. McMellon, C; Tisdall, EKM. Children and young people’s participation rights: looking backwards and moving forwards. Int. J. Child. Rights; 2020; [DOI: https://dx.doi.org/10.1163/15718182-02801002]

23. Nightingale, B. et al. Co-Production in Service Delivery: Opportunities and Barriers (University of Exeter, 2016).

24. South China Morning Post. More than 16 per cent of Hong Kong’s young people have likely psychiatric problem, large-scale studyfinds. South China Morning Post. South China Morning Post. https://www.scmp.com/news/hong-kong/health-environment/article/3220259/more-16-cent-hong-kongs-young-people-have-likely-psychiatric-problem-large-scale-study-finds (2023).

25. Wong, CS et al. The Hong Kong youth epidemiological study of mental health (HK-YES)—A population‐based psychiatric epidemiology study of youth mental health in Hong kong: A study protocol. Early Interv Psychiatry; 2023; [DOI: https://dx.doi.org/10.1111/eip.13364] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37438914]

26. Zhu, S; Li, X; Wong, PWC. Risk and protective factors in suicidal behaviour among young people in Hong kong: A comparison study between children and adolescents. Psychiatry Res.; 2023; [DOI: https://dx.doi.org/10.1016/j.psychres.2023.115059] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38150811][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10332913]

27. Bagnall, CL; Skipper, Y; Fox, CL. Primary-secondary school transition under Covid‐19: exploring the perceptions and experiences of children, parents/guardians, and teachers. Br. J. Educ. Psychol.; 2022; [DOI: https://dx.doi.org/10.1111/bjep.12485] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35132610][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9111429]

28. Lai, KYC et al. School-Based mental health initiative: potentials and challenges for child and adolescent mental health. Front. Psychiatry; 2022; [DOI: https://dx.doi.org/10.3389/fpsyt.2022.866323] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36699476][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9764334]

29. Legislative Council Secretariat. Mental health of local students. (2024). https://app7.legco.gov.hk/rpdb/en/uploads/2024/ISSH/ISSH22202420241028_en.pdf

30. Sum, MY et al. Stigma towards mental illness, resilience, and help-seeking behaviours in undergraduate students in Hong Kong. Early Interv Psychiatry; 2024; [DOI: https://dx.doi.org/10.1111/eip.13455] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37438914]

31. Lai, AHY et al. Life-World design: A career counseling program for future orientations of school students. Child. Youth Serv. Rev.; 2024; [DOI: https://dx.doi.org/10.1016/j.childyouth.2024.107627]

32. Education Bureau. Student Mental Health Information Online - home. (2023). https://mentalhealth.edb.gov.hk/en/

33. Education Bureau. Student Mental Health Information Online - Policy Highlights. (2021). https://mentalhealth.edb.gov.hk/en/whole-school-approach/policy-highlights.html

34. Thiang, O et al. Mental health promotion in Hong kong: A service evaluation of the coolminds pilot study. J. Sch. Health; 2024; [DOI: https://dx.doi.org/10.1111/josh.13492] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39054852]

35. Docksey, AE et al. The stigma and Self-Stigma scales for attitudes to mental health problems: psychometric properties and its relationship to mental health problems and absenteeism. Health Psychol. Res.; 2022; [DOI: https://dx.doi.org/10.52965/001c.35630] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35928586][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9346955]

36. Neff, KD et al. Development and validation of the Self-Compassion scale for youth. J. Pers. Assess.; 2021; [DOI: https://dx.doi.org/10.1080/00223891.2020.1729774] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32125190]

37. Torsheim, T; Wold, B; Samdal, O. The teacher and classmate support scale: factor structure, test-retest reliability and validity in samples of 13-and 15-year-old adolescents. Sch. Psychol. Int.; 2000; [DOI: https://dx.doi.org/10.1177/0143034300212006]

38. OECD. PISA 2018 results (III): what school life means for students’ lives. OECD Publishing; 2019; 2019eedl.book...O [DOI: https://dx.doi.org/10.1787/acd78851-en]

39. Callahan, JL et al. Enhancing information pertaining to client characteristics to facilitate evidence-based practice. J. Clin. Psychol.; 2013; [DOI: https://dx.doi.org/10.1002/jclp.21995] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23801517]

40. Gaiha, SM et al. Effectiveness of arts interventions to reduce mental-health-related stigma among youth: a systematic review and meta-analysis. BMC Psychiatry; 2021; 21, pp. 1-26. [DOI: https://dx.doi.org/10.1186/s12888-021-03350-8]