4.1. What Barriers Do Korean Teachers Face in Implementing ESD?
I conducted a descriptive analysis to examine the distribution for single variables before running the LCA. Barriers to implementation of ESD consisted of fourteen different statements that teachers may face as barriers. Teachers’ response percentages on barriers to implementing ESD can be seen in Figure 1
As can be seen in Figure 1
, there were considerable perceived barriers to ESD implementation to the extent that: approximately 90% of Korean secondary teachers surveyed said they have a lack of information (89.5%), lack of instructional materials (89.1%), lack of knowledge about SD (87.1%), and lack of PK in ESD (87.1%). These perceived barriers were considered to exist with regard to comprehension and understanding of the concept and their teaching skills, and where teachers can access materials to help them implement ESD in their teaching practices, which belong to the individual barriers’ component. Furthermore, there was considerable lack of interest (74.5%). Amongst Korean secondary teachers are confused about concept of SD (75.9%) and have lack of interest (74.5%).
For perceived structural barriers, Korean teachers responded as follows: too much school administrative work (82.2%), curriculum too overcrowded (73.4%), lack of funding (68.0%), class size too large (58.0%), not included in one’s teaching contents (55.2%), lack of incentives (45.0%), and lack of principle support (44.3%).
Therefore, Korean teachers surveyed were more likely to perceive individual barriers than structural ones. Taken together, most teachers surveyed perceived many obstacles in the classroom: insufficient competency, lack of support or resources, large class sizes, and time constraints.
4.2. What Types of Perceived Barriers Exist among Korean Teachers?
To identify if there was a latent class structure that adequately represented the heterogeneity in perceived barriers to ESD among Korean teachers, and if so, what types were they, I carried out an LCA. Since two components were identified (Table 3
), the underlying assumption in each was guided either by individual barriers, structural barriers, or the combination of the two, therefore, generating four possibilities. I repeated the use of PROC LCA for 2–4 class models each with 14 barrier indicators, and the point was to explore possible baseline models. Model fit statistics for LCA models with different numbers of latent classes are shown in Table 4
As can be seen in Table 4
, while the AIC values were the lowest, BICs were slightly larger in the four-class model compared to other classes; however, taken together with other statistics, the four-class basic model was considered as optimal. This is because the four-class basic model had the highest entropy of 0.82 with a better interpretation compared to others, also indicating a good quality of latent class classification.
In addition to the technical analysis, I explored whether the four-class model was interpretable. Specifically, I explored whether each class could be distinguishable from the others on the basis of the item-response probabilities, no class was trivial in size, and it was possible to assign a meaningful label to each class [36
As shown in Table 5
, the four-class model was distinguishable and non-trivial, and meaningful labels could be assigned to each class (Table 5
.) Therefore, the four-class model was selected as the baseline model of barriers to implementing ESD among Korean secondary school teachers.
Each column in Table 5
shows the item-response probabilities for endorsing each item, probability of membership, and the assigned label for each class. Each row represents a different item, and the four columns of numbers are the item-response probabilities of answering “yes” to the item, given that someone belonged to that class. The item response probabilities indicate the probability for an individual to provide a certain response to a specific item given that she or he has been classified in a specific latent class, which show the differences in response patterns that help me distinguish the classes. The probabilities of class membership show the probability for an individual’s membership in a certain class.
Looking at the pattern of responses for all the classes, I obtained an overall picture of the meaning of the four classes, which helped me label them appropriately and meaningfully. Generally, teachers in Class 1 were much less likely to respond “yes” to all variables than the other classes, while teachers in Class 4 were much more likely to respond “yes” to all variables than the other classes.
Those in Class 2 (95%), Class 3 (89%), and Class 4 (100%) had a high probability of saying “yes, lack of PK in ESD”, while only 14% of those in Class 1 respond yes to lack of PK in ESD. In Item 8, 80% of those in Class 4 say they have a lack of incentives. By contrast, those in Class 1 (15%), Class 2 (0%), and Class 3 (31%) had a low probability of saying “yes, lack of incentives” For Item 13, few of those in Class 1 (24%) and Class 2 (42%) have a lack of funding for materials, while a considerable number of those in Class 3 (75%) and Class 4 (86%) said they have a lack of funding for materials.
As shown in Table 5
, Class 3 was similar to Class 4 on nine items (1, 2, 3, 5, 6, 10, 11, 12, and 13). But for Item 8, 80% of those in Class 4 said they have a lack of incentive, while only 30% of Class 3 said they have a lack of incentive. In addition, for Item 14, 90% of those in Class 4 said that ESD was not included in one’s teaching contents, while only 20% of Class 3 said that ESD was not included in one’s teaching contents. Focusing just on each class, the characteristics of each class are as follows:
Class 1 (the Few barrier): Less than 50% of teachers in Class 1 were likely to respond “yes” to all variables. Few of them had a lack of knowledge about SD (35%), few had a lack of PK in ESD (14%), and few were confused regarding SD (16%). Especially, for lack of competency (0%) and lack of principle support (5%), they rarely answered “yes”. It seems that those in Class 1 reported very few barriers. Therefore, this class was named the “few barrier” type. Only 10% of teachers belonged to the few barrier type.
Class 2 (the Individual barrier): Those in Class 2 had a high probability of saying “yes” to the individual component, while they had a low probability of saying “yes” to structural components. For Item 2 and 5, the probability of answering “yes” to each question was 100%. By contrast, the probability of answering “yes” to Items 10 and 12 was only 20%. Furthermore, for Items 8 and 9, the probability of answering “yes” to each question was 0%. Therefore, this class was labeled the “individual barrier” type. Approximately 20% of teachers belonged to this class
Class 3 (the Combination of Individual and Class-Driven Structural barrier): Those in Class 3 had a high probability of saying “yes” to the individual components. Those in Class 3 had a high probability of saying “yes” to some of the structural barriers (10, 11, 12, and 13), while they had a low probability of saying “yes” to other structural barriers (8, 9, and 14). Class size, administrative work, and too overcrowded curriculum are related to class circumstances. A lack of funding for materials can be directly related to class-driven barriers of structural barriers. On the other hand, non-inclusion in one’s teaching contents is related to the subject’s syllabus in compulsory schools, and incentives and principal support can be relevant to institutional practices and policies. As a result, lack of incentives, lack of principle support, and non-inclusion in one’s teaching contents can be related to institutional barriers. Based on these characteristics, this class was termed the “combination of individual and class-driven structural” barrier type. Almost 30% of total respondents belonged to this type.
Class 4 (the Combination of Individual and Structural barrier): Those in Class 4 have lack of PK in ESD (100%), lack of knowledge about SD (97%), lack of information (98%), lack of instructional materials (95%), lack of awareness (93%), lack of interest (88%), lack of funding for materials (86%), and lack of incentives (80%). Also, a very large number said they have too much administrative work (100%), were confused regarding SD (98%), that ESD is irrelevant to their subject (90%), and their curriculum was too overcrowded to add ESD (84%). Seventy percent of those in Class 1 had a lack of principal support and suffered from class sizes that were too large (76%). Therefore, this class was named the “combination of individual and structural” barrier type. This type accounted for approximately as much as 43% of the teachers.
To provide a visible summary of the four-class model, represented as four different lines, the profile plot of the four-class model is given in Figure 2