PROCEDURAL E-SCAFFOLDING IN IMPROVING STUDENTS’ PHYSICS PROBLEM SOLVING SKILLS

Article Info ________________  Received April 2018 Accepted June 2018 Published July 2018  ________________


INTRODUCTION
Problem Solving Skills (PSS) is an important domain in supporting "life capabilities" to lead to a knowledge era "21st century" (Antonenko et al., 2014;National Research Concuil, 2010;Trilling & Fadel, 2009). In the context of education in Indonesia, PSS is adjusted to the curriculum at each level of education as stipulated in the Indonesian National Qualification Framework. PSS not only makes individuals able to solve mathematical problems but also makes them able to solve problems in the surrounding environment (Agustina et al., 2018;Praptiwi et al., 2018;Hertiavi et al., 2016). PSS is not enough if it is only owned by each individual, but it needs a routine to be honed and improved because the problems that will be encountered in the environment tend to have different levels of difficulty.
Especially in physics learning, there is always a session where PSS is honed after students receive learning material, especially quantitative physics problems (McDaniel et al., 2016;Lin, 2015). Students often have difficulty in solving physics problems so that it becomes an obstacle to improving their PSS. In line with this interpretation, a number of studies show several reasons that make physics PSS difficult to improve because it requires complex mathematics (Prahani et al., 2016), requires high metacognition (Saputri & Wilujeng, 2017), often occur misconceptions (Halim, et al., 2014;Alwan, 2011) and lost the concepts (Von Aufschnaiter & Rogge, 2010).
Following up on the explanation of the reasons above, that in the process students will usually be able to solve problems when presented by familiar problems. But if faced with an unfamiliar problem they will have difficulty, this is because the problem is outside their actual ability so that scaffolding is needed to help it solve the problems.
Scaffolding is social supports (Cheng et al., 2018;Belland, 2017;Amelia et al., 2016;Vygotsky, 1978) provided by individuals who have excess experience to individuals who have less experience. Scaffolding activity is usually held by peer interaction in small groups. Social supports makes individuals out of their actual zones leading to the zone of proximal development (ZPD). The term of ZPD was introduced by vygotsky (1978) where this zone is the distance between zones where individuals are able to do a task independently and zones where they need the help by competent peers in carrying out their tasks.
Scaffolding is proven to bring students to ZPD (Harland, 2003) and becomes an important attribute of PSS (Tawfik et al., 2018), but there are shortcomings that are still in the spotlight, scaffolding which is often used still in conventional contexts so that students do not routinely interact. For this reason, scaffolding is needed which can be used anywhere and anytime, one of the solutions to this deficiency is to integrate it with technology (ICT). Actually there have been several studies (Chen, 2014;Malayao & Tagupa, 2014;Deejring, 2014) related to scaffolding that is integrated with ICT.
Chen (2014) built the system using adaptive scaffolding that supports the cognitive needs and motivation of students. The results of the research show that the advantages of an adaptive scaffolding system appeal to students and have proven to be able to improve performance improvements and motivation.
Malayao & Tagupa (2014) built the system by mixing ICT and non-ICT scaffolding, their research focus on conceptual scaffolding. The results of their study showed that the groups that used scaffolding were significantly superior to those who did not use scaffolding in their learning activities. Deejring (2014) built his system with a complete types of scaffolding, including conceptual, procedural, strategy and metacognitive. The results of his research, the e-scaffolding system proved to be able to encourage students to build knowledge and improve their competence and support collaborative learning.
From the above findings, researchers are interested in developing e-scaffolding that focuses on procedural types that can be used anytime and anywhere (assisted by online website) to help students solve physics problems. The main characteristic of this procedural e-scaffolding is that remains in a social constructivist scope, can be adapted to students' actual abilities and can facilitate collaboration both synchronous and asynchronous. In terms of the system, The difference of e-scaffolding from the existing escaffolding is that e-scaffolding developed is able to split the solving categories of students who are expert and novice through recording the problem solving process activities and able to provide formative assessment through feedback on the students physics problem solving process. Borg & Gall's R&D research design (1996) in product design stage, expert and practitioner validation, limited implementation and effectiveness test to improve physics PSS. The design of the developed e-scaffolding I was evaluated by experts and practitioners who were competent in the application of technology in physics learning. Expert and practitioner evaluation data obtained were quantitative data and testimonials through questionnaires related to their views on the design of e-scaffolding. After the first draft has been evaluated, it would be revised so as to produce draft II e-scaffolding that would be tested on students for their readability. The data retrieval technique was by collecting empirical findings that appear in the limited implementation of draft II and will be reviewed descriptively, which then becomes a reference for revising draft II to become a draft III of e-scaffolding.

This study is
The next stage is testing the effectiveness by using quantitative research methods where two classes are taken from the population in a cluster random selection, one class is used as an experimental class (N = 31) and the other class as the control class (N = 32). The experimental class using design III e-scaffolding that used blended learning while in the control class using conventional learning. The physics learning materials used at this stage was particle dynamics. The research design uses pre-post control group design. The technique of retrieving the data through problem solving tests with essay questions before and after treatment is then analyzed to obtain the increase value. Then, the improvement value is analyzed by T-test to see whether there is a difference between learning in the experimental class and the control class.

RESULTS AND DISCUSSION
The problem used in this study was a nonroutine problems (Figure 1). Physics information in the problems was summarized in the form of visual representation and was not verbally told with the aim of making students challenged to think analytically to describe precisely the context of the problems.

Figure 1. Example of Non-Routine Problems
E-scaffolding developed was a procedural type in the form of prompt questions that contained questions about Polya's problem solving procedures (1957). There are two types of e-scaffolding that are developed, although both are in the prompt questions but the difference between them is that in e-scaffolding 2 there is specific information on each question. The reason for making these two types of e-scaffolding is that researchers can split students who have PSS in the expert category (accessing escaffolding 1) and novice (accessing e-scaffolding 2 after e-scaffolding 1 has no effect). The characteristics of e-scaffolding developed can be seen in Table 1.
The procedure for using e-scaffolding, when students access e-scaffolding 1, a pop-up would appear which contains 4 prompt questions (what is known ?, what is the problem ?, how is the problem solving strategy ?, how is the result?). After receiving the four questions, students reflect on the their scheme of knowledge on the problems of particle dynamics given. Then, peer scaffolding is perform by asking for direction from his group friends or from other groups who are more competent to complete his knowledge scheme. If the peer scaffolding activity is claimed to have met a meaningful solution in solving particle dynamics problems, students can directly answer the problems, but if there is solution on the problem solving solution or the solution they input is wrong, students can accesses e-scaffolding 2. prompt questions are answered based on peer scaffolding between students where students are more competent to assist their friends in answering the prompt questions Choose one specific information that is considered correct that is displayed at each prompt question. Then revise the choice based on feedback from prompt questions if something goes wrong. When to use scaffolding?
When the problem is outside the student's actual development zone When a solution is inputted on a wrong question or when e-scaffolding 1 does not provide a meaningful solution in constructing students' ability to solve problems. Scaffolding description of the problem solving shown in Figure 1.

What is known?
What is the problem?
What is the problem solving strategy?

How is the result?
What is known? • The objects that are pulled up are on the floor then there is a force that presses to form an angle with a horizontal line • The objects that are pulled up are on the floor then there is a force that presses perpendicular to the vertical line

What is the problem? • Normal cuboids force • Weight of cuboids
What is the problem solving strategy? • Determine the normal force of the cuboids by studying the forces acting on the cuboids through the law of Newton 1 • Determine the weight of the cuboids by studying the forces acting on the cuboids through the law of Newton 1 How is the result? (N = 36 N) (Note: green is the right option) E-scaffolding 2 contains prompt questions that are the same as in e-scaffolding 1, but in escaffolding 2 there are aspects that contain correct and wrong instructions at each of the prompt questions and students must choose one of the instructions that they consider correct. When they finish choosing, there will be feedback (formative assessment by the system) of their performance which will help them to diagnose their mistakes at prompt questions. But if the feedback results describe each of the instructions selected are correct at the prompt questions, students are ready to answer the problems.

Blended Learning Method
In a limited implementation, the blended learning method is set to support cooperative learning. We use a method in which online and face-to-face sessions are carried out simultaneously so as to make these two sessions a non-separate entity. The main reason why the use of blended learning methods simultaneously between online and face-to-face is because we want the developed e-scaffolding still in the social constructivist scope.
Online sessions, students open particle dynamics learning materials in website with computer to access particle dynamics problems. Students are also allowed to use other electronic equipment such as mobile phones or smart phones to support additional information search in solving the problems. Face-to-face sessions, where students give or receive cooperative peer scaffolding with their group friends to follow up on the steps they take in an online session.

Evaluation Results of E-Scaffolding by Experts and Practitioner
E-scaffolding feasibility evaluation data was collected after validated by 2 experts (physics lecturers) and 4 practitioners (high school physics teachers) who experienced applying technology in physics learning. Based on the results of the feasibility evaluation by experts and practitioners shown in Table 2, escaffolding was developed in the "very good" category so that it was feasible to be used in classroom learning with some improvements. Testimony from the two experts together in responding to the example of the problem used is physics information from the problem must be delivered with verbal to clarify the contextual of the problem. Then the testimonials from practitioners that became an important note for researchers in revising the e-scaffolding II design are the problems raised above the cognitive level of high school students, so that it is necessary to adjust the problems given.

Limited Implementation of E-Scaffolding
There were several findings related to the limited implementation of e-scaffolding products in the class. These findings were testimony of students and findings that were seen when directly observed when students used e-scaffolding products. These findings are first, peer scaffolding did not occur when accessing e-scaffolding 1 but since students read the problem. It was very apparent that students had started to discuss with their group friends when they read the problem together, even though the researchers hoped that this peer scaffolding session would be when following up on e-scaffolding I that had been accessed. This finding provided a reference that the problem that was accessed whether outside the student's actual zone was not only proven when students answer incorrectly or directly access e-scaffolding I, but also could be known when the peer scaffolding occured when reading a problem.
The second finding, the peer scaffolding activity that occurs was collaborative and not cooperative. Researcher's expectation, peer scaffolding was cooperative activity where students who had excess ability to provide assistance to friends who have less ability because they have more learning experience than their friends who have less ability (Cahill et al., 2018). However, the fact was that students collaborated and covered up each other's deficiencies where they shared their tasks and shared their thoughts and together construct their knowledge to solve problems, this was similar to what was stated by Chang et al. (2017); Shin et al., (2017);and Belland (2010) in his research that students could construct their knowledge through collaborative activities. In addition, collaborative activities can provide cognitive benefits (Weaver et al., 2018). This finding indicated that the schemes of students' knowledge when solving problems were still very minimal. As an important note also that escaffolding developed was still not flexible or in other words it was not really adjusted to the actual level of students. However, through peer scaffolding could cover these shortcomings and e-scaffolding 1 also appeared to play an important role in minds-on activities among students.
The third finding, in the problem solving process there were several students who went to other groups to collaborate because their group friends could not be invited to work together and the e-scaffolding accessed was not very helpful. So that for product development online discussion forums are created to facilitate collaboration between students of different groups without having to move places. Based on this discovery, the next stage of the research (effectiveness test) was used blended learning that supports the collaboration between students both synchronous and asynchronously which were carried out simultaneously.
The fourth finding, 100% of students prefer to look for additional information to solve problems through electronic devices (such as devices) and no one opens the book. This was because getting information would be faster if through electronic media because only by entering the keyword information on the search engine the desired information would be immediately available. Searching through printed books can take a lot of time because they have to look for by reading. It can even cause boredom if they have read a lot but the information they want was not obtained. Although information retrieval via the internet was more efficient, but there were things that become susceptible to the process of information absorption by students, that they are free to access information available on the internet so that a high probability of misconceptions would be found. So, to follow up on these findings, the researchers would include material content of particle dynamics on the validated page of the concept's truth.
Of the four findings on the limited implementation described above, there are actually other problems that are found but only limited to technical problems. Students were confused in carrying out the instructions in the guidelines and procedures for using e-scaffolding developed. Therefore, at the next stage of the research, a problem sample was needed to be simulated as an habituation before accessing the problem with real learning. All of these findings would be followed up to become a reference for the revision of the escaffolding draft II to produce the e-scaffolding design III, which would be tested to increase student physics PSS.

Effectiveness Test
The effectiveness test was carried out for 4 meetings to see the effectiveness of e-scaffolding to improve student physics PSS. The results of student PSS improvement can be seen in Table 3, the improvement in student PSS in the experimental class is in the high category, while in the control class is in the medium category. Then inferential analysis is done to see whether there is a significant difference in the increase in physics PSS between the experimental class and the control class (Table 4). Normality test shows that the number of significance in Shapiro-Wilk is above 0.05, so that the data on increasing physics PSS of students in both classes is normally distributed. The homogeneity test shows that the significance number is also above 0.05 so that the data on increasing the physics PSS of students in both classes has homogeneous variance.
After the test conditions for normality and homogeneity were met, a T-test was carried out. The results of the t-test obtained the value of p < 0,05 so it can be concluded that there was a difference in the increase in physics PSS of students in the experimental class with the control class. Based on the description in Table 3, students in the experimental class who learned using e-scaffolding III design had an increase in physics PSS better than students in conventional learning control classes. The findings on the effectiveness test of escaffolding in the experimental class are not so much different from the findings on the limited implementation. However, the experience at the readability test can suppress the technical and operational constraints that occur during this effectiveness test. Learning activities carried out using blended learning also do not encounter significant obstacles, students were collected in small groups of 2 to 3 people to directly solve problems presented through the particle dynamics website. The task of the teacher as the "control holder" of the learning process (Nurita et al., 2017) and not to deliver learning material, but students were directed to construct their own knowledge by accessing abundant learning resources both in books and on the internet or through collaboration with friends the group. In the control class, learning activities began with introductory materials. Then, they are presented with practice questions, but the practice questions given were not through the website.

Figure 2. Asynchronous Collaboration through Live Chat
The problem solving process by students in the experimental class took place in synchronous and asynchronous collaboration through the live chat column (Figure 2). To find additional information in solving problems, students used teaching materials provided on the particle dynamics website and there were also through e-books they get from surfing the internet. Students were also seen preparing several alternative solutions if their solutions are declared wrong by the feedback system.
The problem solving process in the control class did not occur peer scaffolding activities, but it appears that students who have excessive experience intervence in other students by providing a solution without explaining how the process of solving it. Then the feedback provided by the teacher regarding the results of problem solving is not followed up by the students. The e-scaffolding developed was able to split the categories of students who had PSS expert ( Figure 3a) and novice (Figure 3b) through the recording of student performance when solving problems at Figure 1. Based on the recording of problem solving performance in Figure 3a, students with expert problem solving categories were able express the concept of the problem and can qualitatively argue the problem solving process (Riantoni et al, 2017;Sujarwanto et al., 2014;Mason & Singh, 2011), whereas for the novice category, student knowledge organizations tend to be "fragile" (Docktor et al, 2016;Docktor & Mestre, 2014) so that giving specific information at each prompt question is able to strengthen knowledge that is proven by the correctness of all the specific information selected in e-scaffolding 2 (see Figure  3b). E-scaffolding that was developed was able to make students physics PSS gradually improve because the effectiveness test was obtained by recording the performance of students who rarely access e-scaffolding 2 at the last meeting.
Based on the explanations related to the implications of using e-scaffolding above, there are some that are lacking. First, the problem presented is still not contextual with the students' initial knowledge. Secondly, learning activities are limited to minds-on among students in solving problems and there is no hands-on activity so that solution of the problem are less precise.

CONCLUSION
Based on the explanation of the results and discussion, it can be concluded that procedural e-scaffolding has been successfully developed. E-scaffolding is declared feasible to be used and effective for improving student physics PSS.
As a follow-up to further research related to the development of e-scaffolding. First, we hope that the technical page is identified with the computerbased national examination system because it can be a vehicle for habituation of students from an early age to face it. Second, the physics problems presented is more contextual with everyday life.

ACKNOWLEDGMENT
Researchers would like to thank Mr. Supriyono Koes Handayanto and Mr. Sunaryono for involving researchers in the research umbrella. Thanks also to the evaluators of e-scaffolding development, Mr. Sentot Kusairi and Mr. Amiruddin Takda who have provided meaningful advice regarding the products we have developed for the better. My gratitude is also to the entire academic community of Kolaka State High School 1 that has supported and facilitated researchers during the implementation of this research.
(2014). Fostering collaborative problem solving and 21st century skills using the DEEPER scaffolding framework. Journal of College Science Teaching, 43 (6)