The Effect of Discovery Learning Model on Students’ Ability in Determining Character Education Values in Folklore at The Tenth-Grade Students of SMA Negeri 3 Ambon, Indonesia

This research is by a research grant Abstract This research is aimed at finding out the effect of the discovery learning model on students’ ability in determining character education values in folklore at the tenth-grade students of SMA Negeri 3 Ambon. This belongs to experimental research. This research data is the students’ test results in determining character education values in folklore. The number of the sample is 38 students (the experimental class) and 35 students (the control class). SPSS version 23 is used to analyze the data. The result shows that there is an effect of the discovery learning model on students’ ability in determining character education values in folklore at the tenth-grade students of SMA Negeri 3 Ambon. It is proved through the result of the independent sample t-test which is t-count > t-table (4.353 > 1.994).


Background of The Study
Education plays an important role in human life. Through education, human life can be formed into a better one. A good education can prepare and equip students with knowledge, attitudes, and skills. Education is not only obtained in formal educational institutions but also non-formal institutions such as within the family and in one's surrounding environment. In other words, education can be obtained anywhere and anytime. (Wibowo, 2013) states that education should be able to provide a place for self-empowerment based on its character, paradigm, and motives. Therefore, in advancing education, character education is an important requirement for today's young generation. It is a means of shaping a person's morals or character for the better one so that they can be free from the bad things.
Ideally, literature learning can significantly contribute to the success of the desired human development. This can be obtained as long as it is carried out with the right approach, i.e. an approach that can stimulate the occurrence of exercise. This study uses one of the genres of oral literature, namely folklore as a means of determining character education values. (Supratno & Darni, 2015) state that national character education through literary works is very important and it is following the current condition of Indonesian society which is experiencing a moral and ethical decline. The deterioration can be seen in one's behavior, attitudes, and speech. Furthermore, it is said that the values in oral literature can be actualized in everyday life and are used as guidelines in family, community, nation, and state (Supratno, 2010).
In each lesson, there are defined criteria to measure the students' ability to understand the material being taught. One of them is the minimum completeness criteria (KKM). The KKM set in the Indonesian language course at SMA Negeri 3 Ambon is 75. Achieving the KKM requires the seriousness of both teachers and students. Teachers make efforts to achieve KKM by maximizing the learning and assessment process. Meanwhile, students are expected to be able to prepare themselves and be proactive when the teaching and learning process is in progress.
One way to achieve these expectations is to apply the discovery learning model. It is an effective and efficient learning model. It helps students detect problems that they will learn on their own. Through the discovery learning model, they can learn to think analytically and think creatively. It has been widely used in research such as those conducted by Kristin & Rahayu (2016), Atikah (2017), Zulastri (2017), Septiani (2018), Kuncoro (2018), and Reinita (2019). It shows that the discovery learning model has a positive influence on students' learning outcomes for all courses at every level of education.

Research Question
The research question is that "Is there an effect of the discovery learning model on students' ability in determining character education values in folklore at the tenth-grade students of SMA Negeri 3 Ambon?" particular. Folklore can also be defined as cultural expressions of the local community through stories that are directly related to various aspects of the social and cultural values of the community. In the past, folklore was passed down from generation to generation orally (Hutomo, 1991).
Folklore is an inseparable part of the culture and history of a nation. Usually, folklore describes various things that happen, such as the occurrence of the universe, place, or an important event. In Indonesian literature, a folk tale is a form of oral folklore. The word folklore itself is an Indonesian translation of English folklore. Folk means collective, namely a group of people who differentiate it from other groups through physical, social, and cultural characteristics, while lore means tradition. Folk is a culture that is passed down from generation to generation, verbally, or through examples accompanied by gestures or reminders (Danandjaja, 1997).
Hamidy (2003) states that the social functions of folklore include entertainment, solace, self-respect, and even as a means of education. In this regard, (Atmazaki, 2007) argues that the social functions of oral literature include: (1) to refresh thoughts and feelings and even strengthen solidarity, (2) can be used to praise the king, people who are considered sacred, authoritative, and even holy, (3) used as a reflection about the life of ancient society or our ancestors and is also used to express the turmoil of the soul.

Discovery Learning Model
The learning model used in the research is discovery learning. It emphasizes the students' ability to find information and understand the concept of learning independently based on their abilities with the teacher's guidance and supervision. According to Budiningsih (2005), discovery learning is a way of learning to understand concepts, meanings, and relationships through an intuitive process to finally conclude. The application of the discovery learning model has stages that must be considered so that learning is following the objectives to be achieved. Syah (2004) states that the stages of the discovery learning model are (1) stimulation, (2) problem statement, (3) data collection, (4) data processing, (5) verification, and (6) generalization.
According to Alma, et al (2010), the steps of applying the dicovery learning model are:

Stimulation
At this stage, students will first be faced with something that can lead to questions to be solved by students. The teacher can start the teaching and learning process by asking questions, suggesting reading books, and other learning activities that lead to problem-solving preparation. So, at this stage a problem will be given that will stimulate the desire of students to carry out an investigation.

Problem Statement
After being given the stimulus, the next step is that students are given the opportunity to identify problems related to the lesson, then students are guided to formulate a hypothesis.

Data Colletion
When exploration takes place, the teacher also provides opportunities for students to collect as much relevant information as possible to prove whether the hypothesis is true or not. Thus, students are given the opportunity to collect (collection) various relevant information, read literature, observe objects, interview with sources, conduct their own trials and so on.

Data Processing
All information from reading results, interviews, observations and so on, must be processed, randomized, classified, tabulated, even if necessary, it is calculated in a certain way and interpreted at a certain level of confidence. At this stage students must process all the data and information that has been obtained.

Verification
At this stage, students are required to carry out careful examinations to prove whether or not the predetermined hypothesis is true with alternative findings, linked to the results of data processing.

Generalization
The generalization stage or drawing conclusions is the process of making conclusions that can be used as general principles and applies to all the same events or problems by taking into account the results of verification or evidence.

Research Design
This research was quantitative. According to Sugiyono (2017), quantitative research data is in the form of numbers and is analyzed by using statistics. This study used a quasi-experimental design method with a nonequivalent control group design. Thus, this study used two classes, namely the experimental class (X MIPA 1), which amounts to 38 students, and the control class (X IPS 3), which amounts to 35 students. The data obtained by observation, interview, and test. The data gathered from the test were used as the main data and were analyzed using SPSS version 23.

Variables
In this study, there are two variables, namely the independent variable and the dependent variable. 1. Independent Variable Independent variable is variable that affect or cause changes in the dependent variable (Sugiyono, 2017). The independent variable in this study is "the application of discovery learning learning models". This variable is usually symbolized or denoted by the variable "X" (Martono, 2010).

Dependent Variable
The dependent variable is the variable that is influenced by or that is the result of the independent variable (Sugiyono, 2017). The dependent variable in this study is "the ability to determining character education values in the folklore at The Tenth-Grade of SMA Negeri 3 Ambon". This variable is usually symbolized or denoted by the variable "Y" (Martono, 2010).

Sample
This study used 2 classes as a sample, namely class X MIPA-1 with 38 students as the experimental class (which applied discovery learning model), and class X IPS-3 with 35 students as the control class (which applied the conventional model). The sample determination of the two classes is based on the level of ability which is equivalent. The samples were determined by using purposive sampling technique.

Data Analysis Technique
In Not valid 0,00 < r ≤ 0,20 Very low validity 0,20 < r ≤ 0,40 Low validity 0,40 < r ≤ 0,60 Medium validity 0,60 < r ≤ 0,80 High validity 0,80 < r ≤ 1,00 Very high validity Testing the level of reliability in this study using the Pearson Correlation Analysis Method in the SPSS version 23 program with menu the analyze > scale > reliability analysis. To be able to determine whether the test instrument is reliable or not, it can be seen in the SPSS cronbach alpha output. Reliability < 0.6 is not good, while 0.7 is acceptable and above 0.8 is good (Priyatno, 2017). Table 2. Instrument Reliability Criteria Reliability Coefficient Interpretative 0,80 ≤ r High degree of reliability 0,40 ≤ r < 0,80 Medium degree of reliability r < 0,40 Low degree of reliability To test the normality, the SPSS version 23 program is used with menu analyze descriptive statistics > explore > plots > normality plots with test continue > OK. To find out whether the research data is normal or not, it can be seen in the SPSS test output of normality of the Shapiro-Walk Table, column sig. If the significance value > 0.05, the data is normally distributed.
To test the homogeneity, the SPSS 23 program is used with menu analyze > compare means > one-way anova > options (statistics) > homogeneity of variance test > continue > OK. The decision is made if the significance value is more than 0.05 (> 0.05) then the variants of the data group are the same. Meanwhile, if the significance value is less than 0.05 (< 0.05), the variants of the data groups are not the same.
To test the hypothesis, the independent samples t test was used to test the difference in the mean of the two groups or independent samples. The menu used in the SPSS version 23 program in this study is Analyze, compare means, independent samples, test, define group, continue, ok. The test criteria based on the significance value is if the significance is more than 0.05 (> 0.05) then H 0 is accepted and if the significance is less than 0.05 (< 0.05) then H 0 is rejected.

Students' Pretest Data in Experimental and Control Classes
In this study, the researchers carried out a pretest in the experimental and the control classes which is aimed at knowing the students' initial abilities in both classes. The pretest results were presented in the following table.  Table 3 showed some data, i.e. the average score in the experimental class was 61.66 with details: a minimum score (44), a maximum score (88), a median (63), and a mode (63). The average score of the control class was 51.37 with details: the minimum score (19), the maximum score (81), the median (56), and the mode (56). A frequency distribution is the grouping of data into several classes. The following is the frequency histogram of the pretest score in experimental and control classes.

Students' Posttest Data in Experimental and Control Classes
The posttest was carried out in experimental and control classes to determine the students' final ability after being given folklore material. The following was a description of the posttest result data.  Table 4 showed some data, i.e. the average score of the experimental class was 72.53 with the following details: a minimum score (50), a maximum score (100), a median (75), and a mode (75). The average score of the control class was 61 with details: a minimum score (44), a maximum score (81), a median (63), and a mode (63). A frequency distribution is a grouping of data into several classes. The following was the frequency histogram of the posttest score in experimental and control classes.

Prerequisite Analysis Test
The prerequisite analysis test in this study was the normality and homogeneity tests. The prerequisite analysis test was conducted to determine whether the research data could be used or not.

5.1.1
Normality Test Based on the research results and the obtained pretest-posttest results of the experimental and the control classes, the normality test was carried out using the Liliefors method in SPSS version 23. If the significance value is > 0.05, the data is normally distributed, and if the significance value is < 0.05, the data is not normally distributed. The normality test result can be seen in the following SPSS output.  Table 5 showed the pretest-posttest scores of the experimental and control classes in a normal distribution because the data showed a significance score of more than 0.05. The experimental class pretest score (0.116 > 0.05), the experimental class posttest score (0.121 > 0.05), the control class pretest score (0.333 > 0.05) and the control class posttest score (0.107 > 0.05).

5.1.2
Homogeneity Test The homogeneity test was carried out using the posttest data which are obtained from the experimental and control classes. This test used SPSS version 23. The decision making is if the significance value is > 0.05, the variants of the data group are the same. If the significance value < 0.05, the variance of the data groups is not the same. The homogeneity test results can be seen in the following SPSS output.  Table 6 showed a significance value of more than 0.05. The result is 0.447 > 0.05. Hypothesis testing can be done if the analysis prerequisite test consisting of normality and homogeneity tests has met these criteria. The research hypothesis was tested using independent samples t-test using SPSS version 23. The menu used was analyze ˃ compare means ˃ independent samples t-test ˃ define group ˃ continue ˃ ok.
Independent samples t-test was conducted to determine the difference between two unpaired or independent samples. The requirement for hypothesis testing is that the data must be normally distributed and homogeneous. This test used the posttest results of the experimental and control classes. The following table showed the results.  Average Score 61 Based on those data, the statistical analysis of hypothesis testing using the independent samples t-test is as follows: H 0 = There is no effect of the discovery learning model on the students' ability in determining the character education values in the folklore. H a = There is an effect of the discovery learning model on the students' ability in determining the character education values in the folklore.

Posttes Analysis
The decision-making criterion for the independent samples t-test is that if the significance value is > 0.05, H 0 is accepted and H a is rejected, and if the significance value is < 0.05, H a is accepted, and H 0 is rejected. If t-count < t-table, H 0 is accepted, and if t-count > t-table, H 0 is rejected. The result of the independent samples t-test on the SPSS version 23 can be seen in the following table.  Table 9 showed a significant value in the Sig. (2-tailed) was 0.000 or the significance value was < 0.05 (0.000 < 0.05). The sample used in this study was 73 consisting of 38 students (the experimental class) and 35 students (the control class). The score of the degrees of freedom (df) can be calculated with the formula n-2. n is the number of samples, so (df) = 73-2 = 71 with an error rate of 5%. The t-table value for (df) is 1.994. The results of the pretest and posttest in the two classes also have differences. The posttest mean score of the experimental class was 72.53, which was higher than the posttest mean score of the control class, which was 61 with a mean difference of 11.53. Besides, the average score of the experimental class was increase as seen from the pretest result of 61.66 and the average score of the posttest result was 72.53. The difference that can be seen was 10.87. Therefore, it can be concluded that there is a positive effect of the discovery learning model on students' abilities in determining the character education values in folklore.
The process of implementing learning about folklore material in two different classes in class X SMA Negeri 3 Ambon. The two classes received different treatment. Class X MIPA-1 as the experimental class received folklore material with the discovery learning model, while class X IPS-3 as the control class used the conventional model. The application of different learning models is very influential during the learning process of folklore material in the two classes.
The application of the discovery learning model makes students in the experimental class very enthusiastic in the learning process. Students become active, have the courage to ask questions, answer every question given, are courageous and able to convey their opinions and build good communication between peers in the group. Discovery Learning learning model makes students find their own information about the problem they want to look for based on their abilities.
The learning process carried out in the control class was very clear that students seemed passive in accepting folklore material. Only two people seem to be active in the learning process. Meanwhile, the other students only chose to remain silent and wrote down the learning material provided.
Before students in the two classes received material about folklore, the researcher gave a pretest to find out their initial abilities. The two classes were given the same material and questions. After the pretest was carried out, the researcher then delivered the folklore material to the two classes with different learning models. After the learning process was carried out, the researcher gave the two classes a posttest. The posttest is given with the aim of knowing the students' final ability after the learning process.
The pretest and posttest results in the two classes also have differences. The posttest mean score of the experimental class was 72.53 higher than the posttest average score in the control class, which was 61 with a difference of 11.53. In addition, there is an increase in the average value in the experimental class as seen from the pretest results of 61.66 and the average value of the posttest results of 72.53 with a difference of 10.87. So, it can be concluded that the learning process using the discovery learning model has a positive influence on students' ability to determine the values of character education in folklore.
The results of the independent samples t test showed that the value of t count > t table was 4.353 > 1.994 and a significance value of < 0.05 (0.000 < 0.05). Based on this calculation, it is proven that H 0 was rejected and H a was accepted. So there is an effect of the Discovery Learning Model on students' abilities.

Conclusion
Based on the research results, the researchers conclude: 1. The results of the posttest in the experimental and the control classes showed a difference, namely 72.53 (experimental class) and 61 (control class). Thus, the two classes had a difference in the average posttest score of 11.53. Besides, the average score of the pretest and posttest results in the experimental class was increase, i.e. 10.87. 2. The results of the normality and the homogeneity tests showed that the data were normally distributed and also homogeneous. Data that were normally distributed showed a significance value > 0.05. The pretest significance value of the experimental class was 0.116 (> 0.05), the posttest significance value of the experimental class was 0.121 (> 0.05), the control class pretest significance value was 0.333 (> 0.05), and the control class posttest was 0.107 (> 0.05). While homogeneous data is evidenced by a significance value of 0.447 > 0.05. 3. The results of the independent samples t-test showed that the significance value was < 0.05 (0.000 < 0.05), while the value of t-count > t-table, i.e. 4.353 > 1.994. 4. The results of the independent samples t-test indicated that the research hypothesis H 0 was rejected and H a was accepted. So, there is an effect of the Discovery Learning Model on students' ability in determining the character education values in folklore.