The participants of this research recruited participants who voluntarily participated by posting contents specifying the conditions for participation in the experiment on the university recruitment announcement and in-campus bulletin board. Sample size calculations were determined using G*Power 3.1 (University of Düsseldorf, Düsseldorf, Germany). Based on an expected effect size of 0.3, power = 0.8, alpha = 0.05, and the required sample size for within between interactions in a repeated measures analysis is 20. Based on these values, the total sample size was adjusted to 23. According to the sample size calculation, twenty-three healthy undergraduates with an average age of 20.26 ± 2.28 years participated in this research. The participants were eight males and 15 females. The criteria for including participant selection are as follows; (1) those who have no restrictions on outside activities or walks, (2) those who can read and understand the given questionnaire and answer sincerely. On the other hand, the criteria for exclusion of participants are as follows; (1) those suffering from severe stress or depression, (2) those abusing cigarettes or alcohol. The participants were informed of the purpose and progress of the research through prior orientation. All participants who wished to participate in the research finally participated in the experiment after filling out the consent form and received coupons worth USD 39.

This research was performed at the Chungbuk National University’s campus forest in Seowon-Gu, Cheongju City, South Korea. The research site is a deck road with a gentle slope and a universal design of about 1.4km is created. The vegetation of site was distributed pitch pine (Pinus rigida), Korean red pine (Pinus densiflora), Oak (Quercus dentata), Korean chestnut (Castanea crenata), dawn redwood (Metasequoia glyptostroboides), black locust (Robinia pseudoacacia). The height and diameter at breast height (DBH) is more than 15 m and 30 cm. During the research, the weather was clear without rain and had an average temperature of 16–20°C.

This research was set up as a randomized 3×3 crossover design to examine the effectiveness of self-guided and guided forest therapy programs on the mental health benefits of university students.

Twenty-three college students were assigned in a random order to three interventions (self-guided forest therapy, guided forest therapy, and daily routine activities) and participated in each intervention over a one-week washout period between interventions, referring to previous clinical studies (Kim et al., 2014; Lee et al., 2021) to eliminate the carryover effect of the intervention.

Specifically, twenty-three subjects were randomly divided into three groups of 7 to 8 to eliminate the order effect. Each participant had a different order of intervention. All participants participated in all interventions once a week for three weeks, and each intervention was conducted over a one-week washout period to eliminate the carryover effect of the intervention. For example, when the research started, the first group participated in a guided forest therapy program run by a forest therapy instructor. The second group participated in a self-guided forest therapy program. Furthermore, the third group implemented their routine activities.

The research was performed from 30 Sept 2022 to 15 Oct 2022. The self-guided and guided forest therapy programs performed same five activities, including stretching, breathing, walking, meditation, and exercise (Fig. 1).

The self-guided forest therapy program was designed to induce a participant to engage in forest therapy activities in the campus forest independently. For this purpose, we set up a sign for forest therapy activities inside the campus forest. The participants spent 60 minutes walking along the trail, watching guideboard-related forest therapy activities, and following the guidelines. The field supervisor only guided the route of the self-guided forest therapy program. Participants also performed a guided forest therapy program consisting of stretching, walking, breathing, meditation, and exercise under the guidance of a forest therapy instructor for 60 minutes. In order to avoid differences in capabilities between forest therapy instructors, all guided forest therapy programs were conducted by the same forest therapy instructor. The two forest therapy programs differed only in the method of delivering forest treatment activities, and the route, exposure time, and activity contents were configured the same (Fig. 2). Daily routine activities without intervention participated in no forest therapy programs, and only psychological tests were conducted. All participants conducted self-guided forest therapy programs, guided forest therapy programs, and non-treatment activities and performed psychological evaluations after the intervention respectively.

Positive and Negative Affect Schedule (PANAS), the Rosenberg Self-esteem Scale (RSES), and the Connectedness to Nature Scale (CNS) were used for psychological evaluation. The PANAS was developed by Watson et al. (1988) to estimate the participant’s positive and negative emotions. The PANAS is a self-reported measurement in that the 20 items consisting of 10 positive and 10 negative questions are scored on a five-point Likert scale (1 = strongly disagree to 5 = strongly agree). The range of positive emotion scores is 10–50 points, and the higher the score, the higher the positive emotion, which means that it represents energy state, complete concentration, and pleasure. This study employed the Korean version of the PANAS (Lee et al., 2003). The PANAS of this research had high reliability (Cronbach’s α of PA = 0.804; Cronbach’s α of NA = 0.884).

The RSES was developed by Rosenberg (1965) and revised by Jon (1974). The RSES was used to assess the participant’s self-esteem levels. This scale is self-reported that the 10 items consisting of five positive and five negative questions are scored on a four-point Likert scale (1 = strongly disagree to 4 = strongly agree). This scale means that the higher the score, the higher the self-esteem. We employed the Korean version of the RSES (Jon, 1974). The RSES of this research had high reliability (Cronbach’s α = 0.743).

The CNS was developed by Mayer and Frantz (2004) to assess the participant’s connectedness to nature level. The CNS is a self-reported measurement that evaluates how much the respondent feels part of nature. This scale has 10 items, and each item has a five-point Likert scale (1 = strongly disagree to 5 = strongly agree). It means that the higher the score, the higher the connectedness to nature. We employ ed the Korean version of the CNS (Gim et al., 2019). The CNS of this research had high reliability (Cronbach’s α = 0.773).

All statistical analyses were conducted using SPSS 18.0 Windows (SPSS, Chicago, IL, USA). Differences between the self-guided forest therapy program, guided forest therapy program, and routine activities were analyzed by One-way repeated measures analysis of variance (One-way RM ANOVA). When significant differences were found in the One-way RM ANOVA, the post-hoc test was performed as a post-hoc test for comparison between interventions using the Bonferroni test. In all statistical analyses, the p-value <0.05 was used as the significance level. Effect size was used Cohen’s d.

Table 1 shows a comparison of the PANAS scores between the three interventions. There was no significant difference in PA (F = 20.505, p = .093), but the effect size was 0.337, which indicates a medium effect size. There was a significant difference in NA (F = 7.588, p = .003), and the effect size was 0.587, which indicates a large effect size. In addition, a post hoc analysis of the NA score difference between the three conditions (Table 1), the guided and self-guided forest therapy program had a significantly lower NA score than the daily routine activities.

This result partially accepted the hypothesis that the self-guided and guided forest therapy program would improve participants’ emotions compared to their daily routine activities. This result showed that the self-guided and the guided forest therapy program improved the negative emotions even though there was no significant change in the participants’ positive emotions. The findings partially correspond with previous studies showing that forest walking activities improve the emotional state (Bratman et al., 2015; Brooks et al., 2017; Takayama et al., 2014). For example, Takayama et al. (2014) examined 45 men on the emotional and resilient effects of walking and viewing at forests and cities for a short time. As a result, it was found that the group who walking or viewing the forest had higher positive emotions, lower negative emotions, and higher resilience and vitality than the group who did the same activities in the city. In addition, Bartman et al. (2015) investiaged 70 young people on the benefits of walking in green spaces and reported that the group walking in green spaces reduced negative emotions and increased positive emotions more than those walking in urban areas. However, previous studies have compared cities and forests with changes in participants’ positive and negative emotions. This indicates that the forest environment has higher positive emotions and lower negative emotions than the urban environment. Since this study is a result of comparing the type of intervention and the presence or absence of intervention, there may be some differences from the results of previous studies.

Table 2 shows a comparison of the CNS scores between the three interventions. There was a significant change in the CNS score (F = 8.262, p = .001), and the effect size was 0.613, which indicates a lagre effect size. In addition, a post hoc analysis of the CNS score difference between the three conditions (Table 2), the guided and self-guided forest therapy program had a significantly higher CNS score than the daily routine activities.

This result accepted the hypothesis that the self-guided and guided forest therapy program would increase participants’ connectedness to nature compared to their daily routine activities. In other words, this result showed that self-guided forest therapy programs and guided forest therapy programs significantly increased the natural connectedness of university students more than daily life activities. This result corresponds with previous studies on natural exposure to natural connectedness (Keenan et al., 2021; Warber et al., 2015). For example, Warber et al. (2015) examined the effectiveness of the four-week nature camp on 36 young adults and reported that it positively improved emotions and social interaction and enhanced participants’ natural connectedness. Keenan et al. (2021) investigated the effects of natural walking on well-being and emotions, including natural connectedness, in 50 adults with depression or anxiety disorder. As a result, when participants performed natural walking, their natural connectedness, and well-being were higher than when they performed urban walking, and their negative feelings were lower. It was also consistent with previous results that natural exposure, regardless of the presence or absence of a group or guide, increased natural connectedness (Lim et al., 2020; Rogerson et al., 2020).

Table 3 shows a comparison of the RSES scores between the three interventions. There was a significant change in the RSES score (F = 7.308, p =.005), and the effect size was 0.576, which indicates a large effect size. In addition, a post hoc analysis of the RSES score difference between the three conditions (Table 3), the guided forest therapy program had a significantly higher RSES score than the routine activities. However, the self-guided forest therapy program also had a higher RSES score than daily routine activities, but there was no significant difference.

This result partially accepted the hypothesis that the self-guided and guided forest therapy program would increase participants’ self-esteem compared to their daily routine activities. In other words, this research showed that the guided forest therapy program increased participants’ self-esteem more than daily life activities. This result is consistent with previous research on forest therapy programs for self-esteem (Lim et al., 2014; Lee et al., 2017; Kim et al., 2020). However, the self-guided forest therapy program showed higher self-esteem than daily life, but there was no statistically significant difference. In the case of the guided forest therapy program, it is believed that the formation of social bonds between participants while engaging forest therapy program had a positive effect on improving self-esteem. This is because social bonds and relationships are known to be closely related to self-esteem. Looking at studies related to social bonding and the relationship between relationships and self-esteem, Harris and Orth (2019) conducted a meta-analysis of the effects on social relationships and self-esteem. As a result, it was reported that social relationships had a significant potential effect on self-esteem. Stinson et al. (2008) analyzed the relationship between social bonds, self-esteem, and health outcomes of university students for ten weeks using a longitudinal design, predicting a lower-quality social bond and a rapid decline in self-esteem. Self-esteem can be understood as a positive or negative evaluation of oneself (Rosenberg et al., 1995). It is known that high self-esteem can effectively interact with the treatment environment and positively cope with high levels of adverse life events (Orth and Luciano, 2015). Therefore, if participating in the forest therapy program improves self-esteem, it may be appropriate to use a guided forest therapy program rather than a self-guided forest therapy program. In this way, guided and self-guided forest therapy programs can be appropriately used according to the purpose of participation.

This study showed that the self-guided forest therapy program could help psychological health like guided forest therapy program. However, these findings do not indicate replacing a guided forest therapy program with a self-guided forest therapy program. Instead, it proposes the self-guided forest therapy program as a new way to supplement it in a situation where there is a limit to implementing a guided forest therapy program. Generally, the guided forest therapy program consists of groups under the guidance of forest therapy instructors. Moreover, the most significant difference in self-guided forest therapy programs is that participants do forest therapy activities on their own without a guide and without being restricted by the group. Therefore, the advantage of self-guided forest therapy is that participants can work in the forest at their own pace because participants are not in a group. So, it can be applied to relieve stress and improve the health of modern people living in busy cities by using urban forests.