From August 21 to November 27, 2021, a GAP was conducted twice a week for a total of 30 sessions: every Wednesday and Saturday for 3 hours per session. After dividing the space on the D University campus into three zones (healing garden, walking trail (dulle-gil), and resting garden), a GAP was conducted in the healing garden, including horticultural practice and monitoring of garden plants, and the subjects were allowed to take a walk and rest on the trail and resting garden. An expert group consisting of horticultural professors, landscape architects, social workers, and gardening activity experts set an effective program direction and formed detailed program activities. The program was conducted by one main facilitator (a doctorate-level student in horticulture) and two assistant facilitators (a Master of horticultural therapy and a student enrolled in the department of landscape architecture). Each session was divided into the introduction part for 30 minutes (stretching, activity guide, tool usage guide, etc.), the activity execution part for 1 hour 20 minutes (garden activities per session), and the closing part for 40 minutes (finishing and organizing tools, reviewing activities, and walking on campus). Interventions in each of the following four areas were set as the basic direction of the GAP: emotional and psychological support, improvement of physical health, provision of knowledge and educational opportunities, and improvement of social skills. On this basis, the program provided the group with emotional and psychological support through relaxation, contact with nature, and opportunities to immerse in gardening activities; helped improve their physical health through exercise and muscle strengthening; provided knowledge about plants and gardening; and enabled them to interact with other participants and form a bond while participating in the program. The program was divided into gardening, events, and theory. Gardening was carried out in five categories: learning about gardens, planting, garden design, gardening records, and garden maintenance. From the end of August to September, the late summer, the participants learned and practiced about plants and soil in the garden as activities of learning about gardens, and they planted plants and created flower beds for an autumn garden. In October, they designed a container garden themselves, recorded gardening activities such as seed harvesting, and learned about garden maintenance. At the end of November, the program was completed by preparing for the wintering of plants. Events were held in October as a picnic and experiences using garden crops, and also included orientation and theory programs. This study focused on the effects of gardening activities with the arrangement of 77% of the total program for gardening (23 sessions in total), 16% for events (5 sessions), and 7% theory (2 sessions; Table 1).

The detailed activities of the program started with gardening activities to understand healing gardens and plants on the campus, and learn basic knowledge about garden maintenance including soil and fertilizer. In Session 5, the subjects planned a vegetable garden themselves in the planter, and planted vegetable seeds and seedlings. In Sessions 6–8, an autumn flower garden was made in the healing garden and plants were planted. In Sessions 9–10, a hub garden was created using herbs and aromatic plants. In Session 11, sprouts and seedlings of the vegetable garden were managed, and in Session 12, compost was added to the garden to supplement the insufficient nutrition. In Session 13, the subjects learned about container gardens and planted Cyclamen persicum in pots. In Session 14, they mulched the ground around the plants with bark and gravel for garden maintenance, and in Session 15, they designed and made a container garden themselves. In Session 16, a flower basket was made with flowers from the autumn flower garden, and in Session 17, garden maintenance was performed, including removing weeds in the garden. In Session 18, the seeds of garden plants were harvested, recorded and stored. In Session 19, smudge sticks were made using herb plants from the herb garden. In Session 20, an autumn garden picnic event was held for participants to take a walk in the gardens and trails on the campus. In Session 21, a guest lecturer was invited to enable the participants to learn about and observe insects, birds, and animals in the garden. In Session 22, they created a rock garden and planted plants that can grow in rock gardens. In Session 23, using cockscomb they planted in Session 6, they learned from an expert how to make cockscomb flower tea, and tasted it. In Session 24, they studied garden plants while walking through the campus, and in Session 25, they created their own Christmas garden in a container. In Session 27, they studied the plants in the garden and made a garden plant book. In Sessions 28 and 29, they prepared for the wintering of garden plants and cleared the garden; participants dug up bulbs, installed wind breaks for shrubs and perennial plants, and covered trees with jute tapes to protect them. In Session 30, a garden party was held to review the activities with the participants and finalize the program (Table 2).

To verify the psychological effects of the gardening program, a parenting stress test, Korean-style mental health assessment (MHA), and satisfaction survey on the program were conducted. The Korean version of the Parenting Stress Index (K-PSI-4) is a reliable and valid measurement tool, which was developed and standardized by translating the latest version of the US Parenting Stress Index (4th edition) (PSI-4, Abidin, 2012), into Korean (Jung et al., 2019; Child and Parent Domains, and total stress score: Cronbach’s α = 0.90–0.97; sub-scales: Cronbach’s α = 0.78–0.87). The K-PSI-4 consists of 101 items to measure parenting stress (PS) in the relationship between parents and children and 19 items to measure life stress (LS). In this study, the results of two domains of PS (Child and Parent domains) were analyzed, excluding the LS domain that is not included in the total PS values. Each item is evaluated by dividing into 6 subscales of Child domain (Distractibility, Adaptability, Reinforces parent, Demandingness, Mood, Acceptability) and 7 subscales of Parent domain (Competence, Isolation, Attachment, Health, Role restriction, Depression, Relationship with spouse). Each subscale in the Child domain is not the children’s actual characteristics, but the characteristics of children as perceived by the parents. It can be said that the four variables of Distractibility, Adaptability, Reinforces parent, and Demandingness are subscales related to children’s disposition, while Acceptability and Mood are subscales that are influenced by parental personality and perception of the self (Jung et al., 2019). First, the subscale Distractibility indicates the degree of hyperactivity, distraction and short attention span, non-conformity, and difficulty in concentrating on tasks; Adaptability represents children’s ability to adapt to changes in the physical and social environment; Reinforces parent indicates the extent to which parents experience positive responses from their children; Demandingness refers to the degree to which children place demands, including through crying, physical clinging, and frequent requests for help; Mood indicates the degree to which children’s emotions such as depression and excessive crying are expressed as dysfunctional behaviors; Acceptability represents the degree to which a child’s physical, mental, and emotional characteristics match the parents’ hopes for their child. In the Parent domain, the subscale of Competence refers to the sense of competence felt toward parenting as a caregiver; Isolation refers to the degree of lack of social support available to fulfill a parenting role; Attachment represents the emotional closeness and bond that parents feel towards their children; Health indicates the current level of the physical health of parents required for parenting behavior; Role restriction refers to the perception of the limitations of freedom and personal identity as a result of parental roles; Depression indicates parents’ depressed emotional state; Relationship with spouse refers to the degree of emotional and behavioral support one receives from one’s spouse to fulfill a parental role. Of the total of 101 items, questions for 91 items were constructed on a 5-point Likert scale ranging from 1 point for “Strongly disagree” to 5 points for “Strongly agree”. For the remaining 10 items, 4 to 5 examples were presented for selection. The sub-domain score was calculated as the total points of the items, and a higher score means higher parenting stress.

A Korean-style MHA was conducted using the online platform MINDEEP, provided by Korea University; five types of test tools were used to measure quality of life, mindfulness, depression, anxiety, and daily vitality. The Korean version of the WHO Quality of Life BREF (WHOQOL-BREF, Cronbach’s α = 0.90) is a tool developed according to the guidelines of the World Health Organization. It is a 26-item instrument consisting of 4 domains related to quality of life: physical health, psychological health, social relationships, and environmental health. The score for each domain is calculated by multiplying the average of the items by 4; the total quality of life score is the sum of all domain scores, ranging from 20 to 100 points; a higher score means a higher quality of life. The Korean Version of the Mindful Attention Awareness Scale (K-MAAS, Cronbach’s α = 0.87) is a scale for evaluating open and receptive awareness and attention, consisting of 15 items on a 6-point Likert scale. It is measured ranging from 15 to 90 points, with higher scores indicating higher levels of mindfulness. The Korean Screening Assessment for Depressive Disorders (K-DEP, Cronbach’s α = 0.94) is a self-reporting test for screening major depressive disorder, consisting of a total of 12 items on a 5-point Likert scale. The Korean Screening Tool for Anxiety disorders (K-ANX, Cronbach’s α = 0.97) is a tool for screening generalized anxiety disorder, consisting of 11 items on a 5-point Likert scale. For K-DEP and K-ANX, to which the item response theory (IRT) is applied to by statisticians, the weight of the selected items is calculated, and the score is calculated by reflecting this weight. For the depression and anxiety scales, higher scores indicate higher levels of depression and anxiety. The Daily Vitality Scale (Cronbach’s α = 0.77) consists of 5 items on a 5-point Likert scale, ranging from 5 to 25 points, with higher scores indicating higher daily vitality.

To evaluate the effect of the program conducted in this study, parenting stress was tested before and after the program using K-PSI-4. A Korean mental health assessment was conducted with the five types of test tools via an online platform before and after the program. A satisfaction survey on the program was conducted after the program using a self-reporting questionnaire consisting of 7 items of satisfaction with overall operation and effect of the program (using mixed response formats including 7-point Likert scale, multiple choice, and narrative response).