AUTOMATION OF HETEROGENEOUS GROUPING FORMATION TO SUPPORT COOPERATIVE LEARNING PROCESS

• Main Author: Nurul Pratiwi, Oktariani
• Format: Dissertations
• Language: Indonesia
• Subjects: Teknik (Rekayasa, enjinering dan kegiatan berkaitan)
• Online Access: https://digilib.itb.ac.id/gdl/view/47359
• Institution: Institut Teknologi Bandung

Cooperative learning is learning carried out together in a group to achieve a common goal. Many studies show that cooperative learning has a great importance on the success of one's learning. One popular cooperative learning method is Jigsaw method. The Jigsaw method requires heterogeneous formations in groups and homogeneous between groups. The traditional grouping process in class usually determined by the teacher, student or randomly. Determination of student group formation conducted by students or random will produce an incorrect formation. Determination of formation by the teacher is difficult because of several factors, the number of students and the attributes. The Jigsaw Method divides students into heterogeneous groups based on attributes of race, ethnicity, ability and gender. However, this attribute is not relevant to the school situation in Indonesia which has almost the same race, Malay. Many heterogeneous grouping studies use different attributes, learning styles, supported subject matter, etc. There are no fixed rules on what student attributes are used in group formation settings. There are many and varied types of student attributes that can be used. The more number of students and attributes used, the more difficult it is for the teacher to determine the group formation manually. Another obstacle is when cooperative learning is held at the beginning of class meetings. In addition, the development of e-learning is also now leading to cooperative learning where it is not possible to determine the group formation process manually, so that this research is needed to help education. The accuracy of formation in a cooperative learning process is very important. The right group formation is able to encourage cooperative learning optimally and be able to increase student understanding as much as possible. Therefore, this study aims to solve the problem of determining optimal group formation for students based on the formation of the Jigsaw method by building a process model that is able to classify students automatically. The Jigsaw group formation is heterogeneous in groups and homogeneous between groups. Determination of group formation in this dissertation based on dissimilarity between students which is calculated using the dissimilarity between variables of mixed type method. The method can be used to calculate the value of inequality between students based on many attributes possessed by students. This method also eliminates student attributes that are of equal value. This makes the system able to accept various types of attributes without limitations, so that this system can be used in the school environment or online learning environment (elearning) under any circumstances (without limitation). The main contribution of this dissertation is the group formation optimization method with the Fixed Root-Optimization algorithm approach. The optimization process of determining the proposed formation begins with the process of determining Fixed Root. Fixed Root is a pair of students who have the highest inequality between students. Fixed-Root is sought as many as the number of groups formed. Fixed-Root is not a way to determine the chairman of cooperative learning in a group. The remaining members of each group were carried out by an optimization process that kept the heterogeneity level in each group high. Random, Switching and Fixed Root-Random grouping method were also performed as comparison algorithms. As a result, the Fixed Root-Optimization algorithm is able to provide heterogeneous fitness values in a fairly high group (0.6) and the fastest grouping time (0.4s). Based on Univariate ANOVA analysis, heterogeneous fitness values are considered very significantly different than other algorithms. Measuring the value of homogeneous fitness between groups is also done. Based on Bartlett's homogeneity test each group produced is considered homogeneous. Complexity of the Fixed Root-Optimization method is O(mn), m is the number of groups and n is the number of students. Student formation generated by the Fixed Root-Optimization method increases the cognitive value of students as well as the cognitive value of students learning in the group formation produced by the teacher. Based on the experimental results and the analysis of the Fixed Root-Optimization algorithm, it is able to form the formation needed by the Jigsaw cooperative learning method.