From March to June 2024, this study was conducted in a junior high school in Minhang District, Shanghai, China. Participants were selected using a random number generator to choose four seventh-grade classes for the study. Two classes were randomly assigned to the experimental group, while the remaining two served as the control group. The intervention was implemented on a class-by-class basis. Students were excluded if they were unable to attend regularly due to reasons such as sick leave, vacations, or personal leave; if they were already participating in other psychological intervention studies; if they chose to withdraw during the study; or if they had a diagnosed mental illness. Both the parents and the students were fully informed about the study, provided consent to participate, and signed informed consent forms. To ensure participant anonymity, each student was assigned a unique random number, known only to them, for the completion of all subsequent evaluations. This study was approved by the Ethics Committee of Shanghai Minhang Mental Health Center (LW202304).

A priori power analysis was conducted using G*Power 3.1 with an effect size of 0.5, an alpha level of 0.05, and a power of 0.8, suggesting a required total sample size of 128 participants (Kang, 2021). Accounting for a potential dropout rate of 15%, the study aimed to recruit 152 participants, with 76 in each group, to ensure adequate sample size.

The intervention course working team consisted of 10 members with diverse expertise: 2 psychology experts who provided guidance on the intervention program’s content, 1 psychiatric nurse expert who offered support for handling special situations during the intervention activities, 4 psychological counselors/psychotherapists who facilitated the intervention activities for the experimental group, 2 psychiatric nurses (not involved in intervention activities) who distributed and collected questionnaires after undergoing standardized training, and 1 graduate nursing student and 1 of the psychiatric nurses who were responsible for organizing, verifying, and entering the questionnaire data.

Before the intervention course commenced, the psychological status of participants in the experimental group was assessed to identify their specific needs for psychological course guidance. The assessment explored their primary objectives for participation, emotional distress, interpersonal relationship challenges, academic difficulties, feelings of meaninglessness, and lack of achievement. Based on the survey results, the course content was tailored and adjusted on a class-by-class basis to better address the identified needs during the subsequent intervention sessions. For example, students who reported high academic stress were provided with engagement-focused sessions aimed at enhancing their involvement in school activities, using strategies such as goal-setting exercises and mindfulness techniques to alleviate stress and improve concentration. In addition, students experiencing interpersonal difficulties received guidance on relationship-building through structured group activities designed to foster communication, empathy, and cooperation. These adjustments were implemented to ensure that each student received relevant support tailored to their individual needs, helping to maximize the effectiveness of the intervention.

The intervention program was based on the five dimensions of PERMA (positive emotion, engagement, relationships, meaning, and accomplishment), and assessments were designed to focus on enhancing these core aspects of well-being. The course consisted of one session per week, each lasting 45 min, for a total of 8 weeks. The control group attended only the standard on-campus psychology curriculum and did not participate in any positive psychology group intervention activities based on the PERMA theory. In contrast, the experimental group received a tailored positive psychology group intervention program structured around the PERMA framework, which included sessions on Introduction, Positive Emotions, Positive Engagement, Positive Relationships, Positive Achievements, Positive Meaning, and a concluding session titled Farewell. The standard program schedule was listed in Supplementary Table S1.

Each session followed a consistent structure, comprising warm-up activities, theme-based exercises, review and summary, and assigned homework. Various teaching strategies were utilized, including group discussions, interactive games, multimedia presentations, question-and-answer segments, scenario interpretations, and case analyses.

The intervention program addressed several key themes. Participants were guided to understand their emotional characteristics and recognize the influence of emotions on behavior and life. They learned to observe others’ feelings, express their emotions appropriately, and employ techniques to maintain positive emotions while effectively releasing negative emotions. Mindfulness exercises were incorporated to enhance focus and awareness of the present moment. The program emphasized identifying sources of happiness within the learning process and developing effective communication skills to improve interactions with parents. Participants assessed their ability to form and maintain friendships and explored strategies to enhance peer relationships. Reflective exercises on significant life events helped participants identify patterns of optimism or pessimism and challenge irrational beliefs to improve self-evaluation. Additionally, the program guided participants in recognizing their personal strengths and setting realistic, stage-appropriate life goals.

This structured and multi-faceted intervention was designed to promote well-being across multiple dimensions, in alignment with the PERMA framework.

At baseline, a general information questionnaire was administered to collect data on class, gender, age, parental relationships, peer relationships, personality traits, life satisfaction, dietary habits, and physical activity levels. Before and after the intervention, three questionnaires were utilized to assess the psychological status of the participants and evaluate the effectiveness of the course intervention.

The Plural Well-being Questionnaire for Middle School Students was developed by a Chinese scholar to assess “feeling good” and “performing well” among middle school students based on the PERMA theory. This questionnaire comprises four dimensions: positive emotions, engagement, accomplishment, and relationships. Each dimension includes four questions, resulting in a total of 16 items. Responses are scored using a 5-point Likert scale, where 1 represents “very inconsistent with me” and 5 represents “very consistent with me.” Scores range from 1 to 5 points for each question, with a total possible score of 16–80 points. Higher scores indicate higher levels of multidimensional well-being. The questionnaire was validated in a sample of 663 junior and senior high school students from Hebei and Shandong provinces. Its internal consistency reliability, measured by Cronbach’s α coefficient, was 0.909, indicating high reliability (Shang, 2014).

The Adolescent Resilience Scale, developed by Hu Yueqin and Gan Yiqun in 2008, assesses resilience across two dimensions: personal strength and support. The personal strength dimension includes three factors: goal focus, emotional control, and positive cognition; while the support dimension encompasses two factors: family support and interpersonal assistance. The scale comprises 27 items, with a total Cronbach’s α coefficient of 0.83, indicating good reliability (Hu and Gan, 2008; Jiao et al., 2022).

The Simple Coping Style Scale, created by Xie Yaning in 1998, is an adaptation of the Ways of Coping Questionnaire developed by Folkman (Zeidner and Endler, 1996). It was designed to reflect the relationship between coping styles and mental health. The scale includes two subscales: positive coping tendency and negative coping tendency, with a total of 20 items. Each item is rated on a four-point scale: “not used” (0 points), “occasionally used” (1 point), “sometimes used” (2 points), and “often used” (3 points). Higher scores in positive coping indicate lower levels of psychological problems and symptoms, while higher scores in negative coping suggest greater psychological difficulties. The scale has a total Cronbach’s α coefficient of 0.90, demonstrating excellent reliability. The coping tendency is calculated using the formula: Coping tendency = Z (positive coping) - Z (negative coping), where Z is the standard score calculated using the mean and standard deviation of the positive and negative coping scores. If the coping tendency score is greater than 0, it indicates a preference for positive coping strategies under stress. Conversely, a score less than 0 suggests a tendency toward negative coping strategies (Xie, 1998).

After data entry by two individuals, statistical analyses were performed using R software (v4.2.2). Normally distributed continuous data were expressed as mean ± standard deviation, while non-normally distributed continuous data were presented as medians with interquartile ranges. Categorical data were described using frequencies and percentages. For between-group comparisons, chi-square tests were used for unordered categorical data. Normally distributed continuous data were analyzed using two independent sample t-tests or analysis of covariance (ANCOVA) with adjustments for covariates. Non-normally distributed continuous data were analyzed using the Wilcoxon rank-sum test. For paired data, a linear mixed-effects model was used, with intervention and potential covariates set as fixed effects and individual IDs treated as random effects.

In addition to analyzing changes in continuous scores of the well-being and resilience scales, we further categorized the scores to examine the proportion of students who reached different threshold levels. Chi-square tests were conducted to compare the proportions of students in the intervention and control groups who scored above specific cutoff points (e.g., ≥10, ≥11, ≥12, etc.) on each variable. This categorical comparison aimed to assess whether the PERMA-based intervention increased the proportion of students reaching higher levels of well-being or psychological resilience, suggesting potential meaningful thresholds for future intervention study (Kunkle et al., 2020; Lipsey et al., 2012). A p-value < 0.05 was considered statistically significant.

A total of 154 students participated in the study, with two classes (n = 77) assigned to the experimental group and two classes (n = 77) to the control group. All students completed the full intervention course and both the pre- and post-intervention assessments. As shown in Table 1, the mean age was 13 years, and gender distribution was similar between groups (48.1% male in the experimental group vs. 55.8% in the control group). Most participants lived with their parents in typical three-person households and reported being satisfied with their parental relationships. Over 90% of the participants reported having at least one close friend, maintaining a normal or healthy diet, and engaging in physical activity at least once per week.

The pre-intervention survey revealed that the primary motivations for participating in psychological counseling were to learn psychological skills (64.3%) and address personal psychological issues (62.9%). The most common challenges reported were high academic pressure or lack of motivation to study, followed by difficulties in interpersonal relationships. To address these concerns, the intervention program included two sessions focused on active engagement, designed to help students improve their concentration and discover ways to derive happiness from the learning process. Additionally, two sessions were dedicated to positive relationships, aimed at improving participants’ interactions with parents and peers.

At baseline, the control group had significantly higher scores than the experimental group for positive emotion (13.8 ± 4.39 vs. 12.4 ± 4.02), relationship (13.8 ± 4.05 vs. 12.4 ± 4.22), and total well-being (51.1 ± 14.4 vs. 47.5 ± 13.6) as shown in Figure 1 and Table 2. After the intervention, no statistically significant differences were observed between the two groups across all five variables.

To further explore the distribution of well-being levels, we compared the proportions of students exceeding key score thresholds. At baseline, a significantly larger proportion of students in the control group scored above thresholds such as 12, 15, and 16 on positive emotion, and 13 or 14 on accomplishment (Supplementary Table S2). However, post-intervention, these differences disappeared. Importantly, the experimental group showed a significant increase in the proportion of students whose total well-being scores exceeded 60 after the intervention (Supplementary Table S3), suggesting that the intervention helped more students reach higher levels of psychological well-being, and this might be a potential threshold for future intervention effects assessment.

Baseline scores on the resilience scale were similar between groups, with no significant differences in total or variable scores (Figure 1; Table 2). After the intervention, the emotional control score in the control group (19.2 ± 3.75) was significantly higher than that in the experimental group (18.0 ± 3.65).

Threshold analysis showed that at baseline, more students in the control group had goal focus scores above 21 and family support scores above 22 (Supplementary Table S4). After the intervention, the control group continued to show higher proportions of students with emotional control above 17 or 18. Meanwhile, the experimental group demonstrated significant improvements in the proportions of students scoring above 15, 18, and 21 on goal focus, above 17 on positive cognition, and above 91 on total resilience (Supplementary Table S5), indicating gains in specific areas of resilience.

Coping tendency scores between the control and experimental groups showed no significant changes before and after the intervention, as shown in Supplementary Table S8. The median coping tendency score in the control group slightly changed from 0.01 [−0.45, 0.33] pre-intervention to −0.01 [−0.29, 0.39] post-intervention (p = 0.905), and the experimental group’s score changed from −0.02 [−0.32, 0.33] to −0.13 [−0.36, 0.35] (p = 0.786). Neither the positive nor negative coping scores changed significantly in either group (all p > 0.05).

In the experimental group, 58 participants had both pre- and post-intervention test results that could be matched using random numbers (paired population). Significant improvements were observed in the scores for positive emotion, accomplishment, engagement, relationship, goal focus, positive cognition, total well-being, and total resilience in this paired population after adjusting for covariates (gender, age, satisfaction with parental relationships, cohabitation status, family structure, number of close friends, dietary habits, and physical activity levels) (Table 3; Figure 2).

However, no significant changes were observed in positive or negative coping tendencies in the paired population following the intervention (Supplementary Table S6).