Conjoint analysis is a typical method used to analyze consumer utility in marketing, predicting the product that will be chosen by the customer by estimating the utility given to each attribute of the product. Conjoint analysis was developed in mathematical psychology and generally refers to the method of mathematically analyzing consumer preferences for multi-attribute stimuli formed by experimental design (Ha, 2012). In conjoint analysis, the subject has a few important attributes, and each attribute is presumed to have a few levels or values. This study conducted a survey on early childhood teachers to implement horticultural activities in the regular curriculum for early childhood education based on the Nuri Curriculum. The survey was conducted online, and demographic characteristics of the subjects and key information of horticultural activities related to early childhood were set considering multiple levels of each attribute for conjoint analysis.

This study set five attributes of horticultural activities for conjoint analysis (Table 1). In the meta-analysis of research papers to verify the effect of horticultural therapy (Jang et al., 2010), the method of horticultural activities was set as growing plants, decorating plants and integrated activities. First, as the method of horticultural activities, growing and decorating plants based on the children’s level was set as the horticultural activity in Attribute 1, and integrated horticultural activities were selected as Level 2, which are activities that integrate other fields like music, art and science with horticultural activities. Attribute 2 is selected based on the types of activities presented in 696 individual activities in 11 teacher’s guidebooks by topic in the 5-year-old Nuri Curriculum (Kim and Woo, 2014, Choi et al., 2020). Types of activities were divided into three levels: big and small group type in which children form groups for activities, free choice type in which children can choose the activity they want by giving them options so that they are provided with educationally meaningful activities, and outdoor play type in which they participate in activities outdoors. Attribute 3 was in three levels of fruit crops, vegetable crops, and floricultural crops as the subjects of horticultural activities. Attribute 4 selected two levels of indoors and outdoors as the venue for horticultural activities. Attribute 5 was in two levels of 30 minutes and 30–60 minutes as the time for horticultural activities.

There were total 72 cases (2×3×3×2×2) of profiles that can be presented according to five key attributes such as method, type, object, place, and time as well as the levels of each attribute in the survey, but evaluation is impossible in reality as it is too much pressure for the respondents to evaluate the preference of all these profiles. Therefore, to adequately reduce the number of stimuli, orthogonal design of SPSS statistical program was used to extract total 20 profiles including 4 profiles for verification (holdout) (Table 2). Profiles for verification are included to verify the validity of prediction by comparing the consistency with actually designed profiles (Heo, 2015).

The part-worth function model was applied to estimate the worth of the survey. The part-worth function model measures utility with the assumption that each attribute and level has individual part worth (Ha, 2012), and it was applied as each level of the attributes of horticultural activities can be presumed to have individual part worth (Fig. 1).

The survey to implement horticultural activities in the regular early childhood curriculum based on the Nuri Curriculum was conducted on early childhood teachers online for 40 days from May 15 to June 25, 2020. 20 profiles were presented for the teachers to prioritize them based on preference, scoring them 10-9 points if most preferred, 8-7 points if preferred, 6-5 points for neutral, 4-3 points for not preferred, and 2-1 points for least preferred. 257 subjects were used in the final analysis out of 320 respondents, excluding 45 with unclear or insincere responses. Conjoint analysis was conducted after statistical treatment according to the rules of the part-worth function model using IBM SPSS v.19.0 program. Correlation among the attributes was verified by Pearson’s correlation coefficient.

The demographic characteristics of the participants are as follows. 90 participants had 4 years or less experience (35.02%), followed by 85 participants with 10 years or more experience (33.07%), 79 with 5–9 years of experience (30.74%), and 3 others (1.17%) (Table 3). Others included interns. 164 participants (63.81%) had at least 5 years of experience as a teacher, which suggests that they might have responded to the questionnaire with general awareness of horticultural activities. By education level, 148 participants graduated a 2-year or 3-year college (57.59%), followed by 77 participants who graduated a 4-year college (29.96%), 21 with a master’s degree (8.17%), others (2.33%), and 5 with a doctoral degree (1.95%). Others included interns currently attending a 2-year or 4-year college. By type of workplace, 126 participants worked in a private kindergarten (49.03%), 118 in a daycare center (45.91%), and 13 in a public kindergarten (5.06%). For teaching certificate, 132 were certified as Teacher Level 2 (51.36%), 124 as Teacher Level 1 (48.25%), and 1 other (0.39%). Other included an intern teacher who does not have a teaching certificate yet. By class age, 121 participants were teaching the 3-year-old class (47.08%), 84 teaching the 5-year-old class (32.68%), 51 teaching the 4-year-old class (19.84%), and 1 other (0.39%). Other included the case in which the class was run as an integrated class for 3-year-olds and 4-year-olds.

The survey respondents were to choose the activity with highest utility calculated as the sum of part worth by attribute level of each profile selected including the profiles for verification (maximum utility rule) (Heo, 2015). Preference of early childhood teachers by attribute of horticultural activities statistically treated according to the part-worth function model is as shown in Table 4. As a result of analyzing preference of early childhood teachers by attribute of horticultural activities, it was found that importance by attribute was object 29.1%, activity type 23.2%, activity method 17.4%, time 16.1%, and place 14.2% (Pearson’s R = .591, p = .008, Table 4). Early childhood teachers showed highest importance in what object is used in horticultural activities among fruits, vegetables and flowers, which indicates that, in designing a horticultural program, it is necessary to first consider selecting the object of horticultural activities, and then consider activity type such as big and small group, free choice, or outdoor play. Preference by level for objects was vegetable crops (part utility[Pu] = .156), fruit crops (Pu = .21), and floricultural crops (Pu = −.177), indicating that the participants preferred vegetable crops most. This study classified herbal plants as horticultural plants. According to Shin et al. (2012), kindergarteners that participated in horticultural activities using herbal plants most preferred growing herbs in pots, followed by making herbal ice cream. It seems it would be desirable to use plants that are ornamental and edible like herbal plants. Preference for place of horticultural activities (14.2%) showed the lowest importance among the attributes, proving that the venue (whether indoors or outdoors) is not a top-priority factor to consider in proposing a horticultural program for preschoolers. As for preference by level of each attribute, the teachers preferred big and small groups for type of activities, and vegetable crops for object of activities. They also preferred indoors for place, and 30 minutes for time.

As a result of analyzing preference by attribute of horticultural activities according to career experience of early childhood teachers, it was found that teachers with 4 years or less experience showed highest importance in object (26.04%), followed by time (25.01%), activity method (21.33%), place (18.75%), and activity type (8.87%) (Table 5). Those with 5–9 years of experience showed highest importance in object (44.25%), followed by time (26.12%), activity method (17.38%), place (10.42%), and activity type (4.98%). Those with at least 10 years of experience showed highest importance in object (51.98%), followed by activity method (22.85%), time (3.91%), place (6.28%), and activity type (4.98%). Others showed highest importance in object (49.12%), followed by place (29.83%), activity type (10.53%) activity method (5.26%), and time (5.26%). The importance was high in object regardless of career experience, and the percentage was 51.98% for at least 10 years of experience, 44.25% for 5–9 years, and 26.04% for less than 4 years, indicating that teachers with more career experience tended to think that selection of the object is more important for preschoolers’ horticultural activities than activity method, type, place, and time. Early childhood teachers with much career experience are also experienced with big and small group activities, free choice activities and outdoor play that are commonly carried out in kindergartens, which is why the importance is measured low, whereas objects such as fruits, vegetables and flowers are considered important in horticultural activities and may change preschoolers’ activities significantly, thereby giving highest importance to object. Importance was highest in object regardless of career experience of early childhood teachers, indicating that they consider object most important in choosing horticultural activities. Moreover, by level of fruits, vegetables and flowers, all of the teachers with 4 years or less experience (Pu = .0593), 5–9 years (Pu = .2089), 10 years or more (Pu = .1907), and others (Pu = .6528) showed higher preference for vegetable crops than fruit and floricultural crops. As a result of analyzing the current state of horticultural education and perception of early childhood teachers, it was found that teachers most preferred edible plants (62.8%), followed by arbors and shrubs (10.5%), and flowering plants (9.3%) (Chi et al., 2018), which is similar to the result that preference and importance of vegetable crops, which are edible herbaceous plants, turned out to be high.

As a result of analyzing preference by attribute of horticultural activities according to education level of early childhood teachers, it was found that teachers who graduated a 2 or 3-year college showed highest preference for object (29.36%), followed by time (28.47%), activity type (23.48%), activity method (16.73%), and place (1.96%) (Table 6). Those who graduated a 4-year college showed highest preference for object (50.40%), followed by activity type (26.75%), activity method (12.39%), time (9.70%), and place (0.76%). Those with a master’s degree showed highest preference for object (51.66%), followed by activity method (31.27%), place (9.47%), activity type (4.75%), and time (2.85%). Those with a doctoral degree showed highest preference for object (32.43%), followed by activity method (22.52%), place (20.72%), activity type (19.82%), and time (4.50%). Others showed highest preference for activity type (32.56%), followed by object (26.75%), place (19.77%), time (17.45%), and activity method (3.48%). All the teachers who graduated a 2 or 3-year college, a 4-year college, have a master’s or doctoral degree except ‘others’ showed highest importance in object. By level, those who graduated a 2 or 3-year college, a 4-year college, have a doctoral degree preferred vegetable crops, while those who have a master’s degree preferred fruit crops. As for activity method, teachers in all education levels preferred exclusive horticultural activities to integrated activities. As a result of analyzing the current state of horticultural education for preschoolers and perception of teachers, it was found that at least a majority of institutions for early childhood education preferred to integrate horticultural education with science or other subjects rather than providing it separately on its own (Chi et. al., 2018), which is different from the result of this study. In this study, teachers preferred to have horticultural activities separately as a single class, which is due to the improved awareness and importance of horticultural activities. All the teachers turned out to prefer exclusive horticultural activities: 2 or 3-year college (Pu = .0650), 4-year college (Pu = .0519), master’s degree (Pu = .1964), and doctoral degree (Pu = .3125). This shows that teachers with a master’s degree and doctoral degree more prefer exclusive horticultural activities than those who graduated a 2 or 3-year college and a 4-year college, showing a difference depending on education level. As for time, teachers with a master’s degree preferred 30–60 minutes, while those who graduated a 2 or 3-year college and a 4-year college, those with a doctoral degree, and others preferred 30 minutes, showing a difference depending on education level.

The Nuri Curriculum is a national-level curriculum comprised of five key areas such as physical exercise/health, communication, social relationship, artistic experience, and nature exploration, designed to promote basic abilities and upright character necessary for 5-year-old children (MEST, 2012). Currently, both kindergartens and daycare centers are adopting the Nuri Curriculum as a regular curriculum. As a result of analyzing preference by attribute of horticultural activities according to workplace of early childhood teachers, it was found that public kindergarten teachers showed highest importance in activity method (35.83%), followed by time (19.17%), activity type (17.50%), place (14.16%), and object (13.33%) (Table 7). Private kindergarten teachers showed highest importance in object (35.39%), followed by time (28.21%), activity method (17.87%), activity type (15.70%), and place (2.83%). As a result of examining the importance perceived by daycare center teachers, they showed highest importance in object (53.44%), followed by time (20.64%), activity method (18.58%), place (4.59%), and activity type (2.75%). All of public kindergarten, private kindergarten, and daycare center teachers preferred exclusive horticultural activities, which is the same case as career experience and education level of the teachers and is related to the increase in the awareness and interest for horticultural activities. However, an additional survey must be conducted to verify the specific reasons. There was no consistency in preference for activity type, object, place, and time among public kindergarten, private kindergarten, and daycare center teachers.

If one has a Teacher Level 2 certificate and at least 3 years of field experience, he or she is qualified to receive Teacher Level 1 training, after which he or she is promoted to Teacher Level 1 (Im and Lee, 2011). As a result of analyzing preference in horticultural activities according to certificate, it was found that Teacher Level 1 showed highest importance in object (46.16%), followed by activity method (31.03%), time (21.73%), place (8.15%), and activity type (2.94%) (Table 8). Teacher Level 2 showed highest importance in object (45.09%), followed by time (20.78%), activity method (19.49%), activity type (9.34%), and place (5.31%). Both Teacher Level 1 and Teacher Level 2 considered object as the most important attribute. By level, the preference was highest for vegetable crops. As for the second-highest preference, Teacher Level 1 preferred activity type, and Teacher Level 2 preferred time, showing a difference between the two.

As a result of analyzing preference of horticultural activities according to the age of children in class which the teachers are in charge of, it was found that the teachers in charge of a 3-year-old class showed highest importance in object (38.83%), followed by activity method (27.56%), place (12.57%), time (12.35%), and activity type (8.66%) (Table 9). Those in charge of a 4-year-old class showed highest importance in time (50.37%), followed by place (20.31%), activity type (8.27%), and activity method (1.50%). Those in charge of a 5-year-old class showed highest importance in object (65.53%), followed by time (16.59%), activity method (9.94%), place (6.15%), and activity type (3.79%). For time, all of the teachers in charge of a 3-year-old class, 4-year-old class, and 5-year-old class preferred 30 minutes. This is a result of considering the time in which preschoolers can concentrate.