Introduction
Agro-healing (also known as care farming, social farming, green care farming, or farming for health) is an industry that utilizes agricultural and rural resources to promote and restore people’s health (Ministry of Government Legislation, 2024). In advanced agricultural countries such as the Netherlands, the concept is referred to by various terms, including care farming, social farming, green care farming, and farming for health. In South Korea, agro-healing is being implemented through farms, rural villages, and institutions (e.g., healthcare, social welfare, and rehabilitation facilities). To support the systematic development of this field, the Act on Research, Development, and Promotion of Healing Agriculture (hereinafter referred to as the Agro-Healing Act) was enacted in March 2020 and came into force in March 2021. This legislation provides institutional support for the development of agro-healing programs for the training of qualified professionals and more.
However, despite the establishment of relevant laws and systems and the development and implementation of various policies, by the end of 2024, there are only about 72 care farms (also known as agro-healing farms) and 17 agrotherapy centers operating nationwide (www.agrohealing.go.kr). This indicates a comparatively slower rate of scale-up and expansion relative to countries such as the Netherlands (1,100 care farms or agrotherapy centers) and France (1,200). Furthermore, public awareness of agro-healing remains low, despite its various effects. One contributing factor appears to be the weak connection between the facilities that implement agro-healing—namely, care farms—and its therapeutic effects. The effects of agro-healing are influenced by numerous factors present in these farms. The more these tangible and intangible factors—such as facilities, resources, operational methods, and programs—are supported and enhanced, the greater the potential therapeutic benefits.
Research on agro-healing can be broadly categorized into two areas: its therapeutic effects and its use and satisfaction. Studies on therapeutic effects include investigations on the therapeutic effects of forest experiences (Song and Lee, 2021), the mediating effects of psychological well-being (Lee and Chung, 2022), the introduction of virtual reality-based physical activity programs (Park et al., 2023), the improvement of cognitive function in older adults with dementia (Kim and Yun, 2024), and the reduction of daily stress (Yeo et al., 2024). Research on use and satisfaction focuses on factors affecting the intention to use agro-healing services (Kim and Ha, 2022), the effects of agro-healing programs on satisfaction and intention to revisit (Ji and Yang, 2023), the effects of such programs on tourism intentions among middle-aged people (Lee and Yang, 2023), satisfaction with agro-healing programs (Oh and Heo, 2021), and preferences for horticultural therapy programs among older adults with dementia (Lee et al., 2024).
Meanwhile, studies on care farms include a survey of experts’ perceptions regarding the design of an agro-healing virtual reality therapy system (Bae et al., 2023), a demand analysis of agro-healing virtual reality therapy system factors based on respondents’ characteristics (Koo et al., 2022), and an analysis of success factors for efficient care farm management (Kim et al., 2024).
However, most of these studies primarily focus on the effects of agro-healing itself, or target only virtual care farms or their managers. Since care farms are the physical spaces where these activities take place and where users directly engage with them, it is essential to establish both tangible and intangible systems that meet users’ needs in order to maximize the therapeutic benefits. Therefore, this study emphasizes that care farms are not merely the settings for agro-healing activities, but are important elements in maximizing therapeutic effects. To this end, we aimed to empirically analyze how tangible factors (such as facilities and resources) and intangible factors (such as operations, programs, and information) of care farms influence the effects of agro-healing and user satisfaction. This study is expected to contribute to policy development by providing a foundation for both tangible and intangible investment in, and support for, spaces where agro-healing is actively practiced. Furthermore, the findings may help promote agro-healing by clarifying the relationship between care farms and the various therapeutic effects and user satisfaction associated with agro-healing.
Research Methods
Variable
Agro-healing refers to activities that utilize resources such as plants, animals, insects, and rural environments and cultures to promote physical, cognitive, psychological, and social well-being; these activities are carried out in agro-healing facilities and target children, adolescents, adults, and the elderly (Ministry of Government Legislation, 2024). A care farm is a place where agro-healing takes place through both tangible factors—such as various facilities and resources—and intangible factors—such as programs, operational capacity, and promotional efforts. Therefore, this study conducted a focus group interview (FGI) with seven agro-healing experts to identify and construct the tangible and intangible factors of care farms; the physical, cognitive, psychological, and social effects of agro-healing; and factors of user satisfaction (Table 1).
Furthermore, 14 items on tangible factors of care farms and 19 items on intangible factors, 27 items on the effects of agro-healing, and 7 items on user satisfaction with care farms were derived as detailed measurement items for each factor. All items were measured using a 5-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree).
The tangible factors of care farms consisted of 7 items related to facilities (Q1–Q7) and 7 items related to resources (Q8–Q14). The intangible factors included 7 items on operations (Q15–Q21), 8 items on program content and implementation (Q22–Q29), and 4 items on information (Q30–Q33). The effects of agro-healing consisted of 8 items on physical effects (Q34–Q41), 4 items on cognitive effects (Q42–Q45), 10 items on psychological effects (Q46–Q55), and 5 items on social effects (Q56–Q60). User satisfaction consisted of 7 items (Q61–Q67; Table 2).
Research model and hypotheses
This study was designed based on the consumption-system approach proposed by Mittal et al. (1999) and systematically analyzed the relationship between care farms and agro-healing effects and user satisfaction, incorporating the effects of consumers’ perceived evaluations on their emotional responses as presented by Bagozzi (1992). Therefore, the study examined the impacts of both tangible and intangible factors of care farms on the effects of agro-healing, as well as how different effects contribute to the overall satisfaction of participants. Additionally, the mediating effects between these variables were analyzed to provide insight into the various effects and to offer implications for the development of programs that could enhance satisfaction with care farms.
To this end, a research model was established as shown in Fig. 1, and the following five hypotheses were proposed:
Hypothesis 1 (H1): Tangible factors of care farms will have a positive impact on physical, cognitive, psychological, and social effects.
Hypothesis 2 (H2): Intangible factors of care farms will have a positive impact on physical, cognitive, psychological, and social effects.
Hypothesis 3 (H3): Physical, cognitive, psychological, and social effects will have a positive impact on user satisfaction.
Hypothesis 4 (H4): Tangible factors of care farms will have a positive effect on user satisfaction.
Hypothesis 5 (H5): Intangible factors of care farms will have a positive effect on user satisfaction.
Data collection and analysis
A total of 600 adult men and women nationwide were surveyed using a stratified sampling methodology based on region, gender, and age. The survey was conducted by a professional online research firm, Macromill Embrain, from June 10 to June 20, 2023. Descriptive statistics for respondent characteristics are presented in Table 3. The mean age of respondents was approximately 44.43 years, with 305 males (50.83%) and 295 females (49.17%). The mean household size was about 2.97 members. Approximately 81.8% of respondents held a college degree or higher, and the most common annual income range was between 20 million and 40 million KRW (24.67%). The collected data were analyzed using the two-step approach proposed by Anderson and Gerbing (1988), as adopted by Lee (2011). Statistical analyses were performed using SPSS 28.0 and AMOS 29.0.
Results and Discussion
Exploratory factor analysis
Prior to conducting an exploratory factor analysis, the Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy and Bartlett’s Test of Sphericity (BTS) were performed to assess the suitability of the data. A KMO value of 0.6 or higher is considered adequate for factor analysis (Kim, 2011), and BTS requires the rejection of the null hypothesis—that the correlation matrix is an identity matrix—in order for factor analysis to be deemed appropriate. Principal component analysis (PCA) was used for factor extraction, and orthogonal rotation using the Varimax method was applied to ensure independence among the factors. The number of factors was determined based on the eigenvalues and the proportion of common variance they accounted for relative to the total variance. Only variables with factor loadings of 0.5 or higher were chosen for further analysis (Kim, 2011).
An exploratory factor analysis of the tangible factors of care farms yielded a KMO value of 0.866 and a BTS result of χ2 = 3,421.002 (p < .001), indicating that the data collected were suitable for factor analysis and that the correlations between variables were statistically significant. Two factors with eigenvalues greater than 1 were extracted, accounting for approximately 52.3% of the total variance (Table 4). In addition, a Cronbach’s α value of 0.7 or higher is generally considered to indicate acceptable reliability (Kim, 2011). The Cronbach’s α values for the tangible factors were all 0.7 or higher, suggesting that they were reliable. Consequentially, the division of tangible factors into two categories—facilities and resources—and the derivation of corresponding measurement items were considered appropriate. Reliability and validity of the factor structure were further assessed through confirmatory factor analysis.
An exploratory factor analysis of the intangible elements of care farms yielded a KMO value of 0.947, and a BTS result of χ2 = 7,207.926 (p < .001). Three factors with eigenvalues greater than 1 were extracted, accounting for approximately 65.8% of the total variance (Table 5). The reliability of each factor was supported by Cronbach’s α values, all exceeding 0.7. Based on these results, the intangible elements were appropriately divided into three categories—operation, program, and information—with corresponding measurement items derived for each. A confirmatory factor analysis (CFA) was subsequently conducted to further assess the reliability and validity of the factor structure.
An exploratory factor analysis of the effects of agro-healing yielded a KMO value of 0.957 and a BTS result of χ2 = 9,001.333 (p < . 001). Four factors—physical, cognitive, psychological, and social effects—accounted for 59.6% of the total variance (Table 6). However, eight items were excluded due to low factor loadings: four items related to physical effects (Q36, Q39, Q40, Q41) and four items related to psychological effects (Q46, Q47, Q53, Q55). These items were removed because their factor loadings were below the threshold of 0.5, which is generally considered the minimum value for a variable to be deemed significant (Kim, 2011). The Cronbach’s α values for each factor—excluding items with loadings below 0.5— were all above 0.7, indicating acceptable reliability. Based on these results, the effects of agro-healing were appropriately classified into four categories: physical, cognitive, psychological, and social, with corresponding measurement items identified for each. A confirmatory factor analysis was then conducted to further assess the reliability and validity of the factor structure.
An exploratory factor analysis on satisfaction yielded a KMO value of 0.916 and a BTS result of χ2 = 2,043.167 (p < .001). A single factor with an eigenvalue greater than 1 explained approximately 60.7% of the total variance (Table 7). The extracted factor had a Cronbach’s α value above 0.7, indicating acceptable reliability.
Confirmatory factor analysis
A confirmatory factor analysis was conducted to verify the reliability and validity of the factors identified in the exploratory factor analysis. Convergent validity refers to the degree to which two measurement methods—as different from each other as possible—are developed to assess the same construct and yet yield highly correlated results. Discriminant validity, on the other hand, indicates that measures of distinct constructs should have low correlations; in other words, when two different concepts are measured, the correlation between the resulting values should be low (Kim, 2011). In this analysis, all standardized factor loadings—except for items Q10 and Q13 under the intangible factor—were above 0.50, supporting the construct validity of the measurement model. Furthermore, reliability and validity are generally considered acceptable when construct reliability is 0.70 or higher and the average variance extracted (AVE) is 0.50 or higher (Kim, 2011). The results of the analysis indicated that most factors exhibited a construct reliability of 0.70 or higher and an AVE of 0.50 or higher, thereby supporting the establishment of convergent validity for the measurement model (Table 8).
To verify discriminant validity, the average variance extracted (AVE) for each factor must be greater than the squared correlation between that factor and any other (Kim, 2011). In this study, the AVE values for each factor generally exceeded the squared correlations with other factors, indicating that discriminant validity is adequately established (Table 9).
To assess the model fit, absolute fit indices and incremental fit indices were estimated based on the χ2 value, CMIN/DF (Q), RMR, RMSEA, TLI, and CFI, as used in previous studies by Lee (2011) and Yang (2015). The model yielded a χ2 value of 3,264.791, a CMIN/DF of 2.637, an RMR of 0.036, an RMSEA of 0.052, a TLI of 0.878, and a CFI of 0.886. These results indicate that the model has a relatively good fit.
Estimation results
Structural equation modeling yielded the following fit indices: χ2 = 1,430.746, CMIN/DF (Q) = 3.415, RMR = 0.032, RMSEA = 0.063, TLI = 0.894, and CFI = 0.904, indicating that the model demonstrated a good fit to the data.
As a result of hypothesis testing, both tangible and intangible factors were found to have a positive (+) impact on the physical, cognitive, psychological, and social effects of agro-healing at the 5% significance level. Therefore, hypotheses 1 (H1) and 2 (H2) were supported (Table 10). Regarding the impact of these four effect dimensions on satisfaction with care farms, only the physical effect showed statistical significance, leading to partial support for Hypothesis 3 (H3). However, Hypotheses 4 (H4) and 5 (H5), which proposed that tangible and intangible factors directly affect satisfaction with agro-healing, were not supported due to a lack of statistical significance. These findings suggest that the physical effect plays an important mediating role in the relationship between tangible and intangible factors of care farms and overall satisfaction (Fig. 2).
Based on these analytical results, several discussion points emerge. First, both the tangible factors (e.g., facilities and resources) and intangible factors (e.g., management, program, and information) of care farms were found to have a positive impact on the four domains of agro-healing effects—physical, cognitive, psychological, and social. This suggests that care farms function not merely as physical spaces, but that the qualitative aspects of program design and operational management significantly influence participants’ therapeutic experiences.
Second, of the effects of agro-healing, only the physical effect had a direct impact on satisfaction with care farms (Fig. 2). This suggests that participants are most responsive to physical changes, probably because they are the most immediately perceptible aspect of care farm programs. In contrast, cognitive, psychological, and social effects may require longer-term engagement to be fully recognized, which may explain why they did not have a significant impact on satisfaction in this study.
Third, the results of this study align with the physical health promoting effects of agro-healing reported in previous research (Park et al., 2023; Lee and Yang, 2023). In contrast, the study by Bae et al. (2023) suggested “psychological and emotional stability” as the most important factor in developing a virtual reality therapy system. This discrepancy likely stems from differences in the research subjects (experts vs. the general public) and the methods used (virtual vs. real-world experiences).
Fourth, these results suggest that public awareness of the non-physical benefits of agro-healing—such as cognitive, psychological, and social effects—remains limited and that opportunities to access related information or experiences may also be insufficient. When individuals have limited exposure to agro-healing or lack specific explanations of its potential effects, it becomes difficult for those effects to translate into satisfaction. Therefore, it is necessary to develop customized programs that sufficiently communicate the multifaceted value of agro-healing while addressing diverse expectations and needs. In addition, efforts should be made to expand continuous participation opportunities and implement education and promotional strategies that bridge the gap between experts and users. Government-level financial and institutional support, along with further research to validate the effects, are also crucial for advancing the field.
Conclusion
This study empirically analyzed the impacts of both tangible and intangible factors of care farms on the effects of agro-healing and user satisfaction, aiming to identify the structural relationships among these variables. The main findings are as follows.
First, exploratory factor analysis derived two sub-factors of tangible factors—facilities and resources—and three sub-factors of intangible factors—operation, program, and information. Confirmatory factor analysis then confirmed the construct, convergent, and discriminant validity of these factors.
Second, structural equation modeling revealed that both tangible and intangible factors positively influenced all four dimensions of agro-healing effects—physical, cognitive, psychological, and social. This indicates that not only the physical environment but also less visible elements, such as program quality, operational management, and information provision, play a key role in shaping users’ therapeutic experiences.
Third, among the four agro-healing effects, only the physical effect had a statistically significant impact on user satisfaction with care farms, whereas the cognitive, psychological, and social effects did not. Neither tangible nor intangible factors directly affected satisfaction; however, both exhibited indirect effects mediated through the physical effect.
These findings suggest that current users of agro-healing programs place the greatest value on physical benefits. In contrast, their expectations for cognitive, psychological, and social benefits appear to be relatively low or may be unrecognized due to limited experience or insufficient information. To promote a more balanced understanding and utilization of the diverse benefits of agro-healing, programs should be designed to incorporate specific information about its cognitive, psychological, and social effects, along with repeated experiential opportunities. It is also necessary to develop customized programs and to implement promotional and educational initiatives to enhance user awareness. These efforts will help ensure a multidimensional understanding of the effects of agro-healing and align its outcomes more closely with the needs and expectations of users.
In addition, the sustainable expansion and activation of agro-healing requires institutional and financial support at the government level. If supported by policy efforts—such as fostering professional personnel, standardizing operational guidelines, and improving program quality—care farms can be firmly established as vital agricultural resources that contribute to improving the quality of life for the public.
This study sought to present implications for the future expansion and activation of care farms, based on the results of a survey on awareness of care farms. With continued monitoring and policy support, it is expected that care farms will continue to evolve through both tangible and intangible expansion to meet user needs; and that a broader population will be able to access and benefit from the diverse advantages these programs offer.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
