Cluster Change Analysis - Assess the Effectiveness of Speaking Skill Techniques Using Machine Learning - with K-Means

Aug 3, 2026

Upload by Sage

Students differ in how they develop speaking skills, so selecting effective instruction methods is essential. This study compares three speaking-teaching approaches using machine learning, examining conventional learning, vlog project learning, and picture series learning. Pre-test and post-test scores from three classes are clustered with K-Means in R to detect how student groups shift after treatment. Results show notable movement among students, with the vlog project class, conventional class, and picture series class showing the largest cluster transitions between pre-test and post-test.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages14

Reading Time35 min

AccessFree to download

Abstract

1 Introduction

2 Material and Methods

Study Area and Research Data

Data Collection

Pre-test and Post-test Clustering

Cluster Verification

3 Results and Discussions

Cluster Change Analysis to Assess the Effectiveness of Speaking Skill Techniques Using Machine Learning Nunik Herani Wulandari 1 , Vega Purwayoga 2* 1 University of Muhammadiyah Tasikmalaya, Indonesia 2 Department of Informatics, Faculty of Engineering, Siliwangi University, Indonesia *Corresponding Author: [vega.purwayoga@unsil.ac.id](vega.purwayoga@unsil.ac.id) (Received 30-08-2024; Revised 22-09-2024; Accepted 07-10-2024) Abstract Students have different characteristics in the learning process, especially in developing speaking skills. So it requires the right learning method. This study aims to compare effective teaching methods for speaking using machine learning. The classes used in this study consisted of three classes: conventional, vlog project, and picture series. The data used were students' pre-test and post-test scores. The machine learning algorithm used is K-Means. KMeans clusters the pre-test and post-test data. The results ofK-Means clustering on the pretest and post-test identified the differences in student groups between the pre-test and posttest. Students who experienced the most cluster movement were those in the vlog project class, the conventional class, and the picture series class. Keywords: Clustering, K-Means, Machine Learning, Speaking Skill 1 Introduction Speaking is one way to communicate with someone. The main components of speaking include fluency, comprehension, grammar, vocabulary, and pronunciation [1] . Speaking skills do not appear instantly, so regular practice is needed. One of the places to practice speaking skills is at school with the help of teachers [2]. Intensive learning can improve speaking skills, especially in English lessons [3] . Based on research [4], students need to develop skills in speaking English. The earlier students learn English, the easier it is for them to understand English. Learning English is an investment for a bright future. Speaking English allows us to easily access and obtain information because Most people write information in English. Some effective teaching techniques for learning to speak include vlog projects and picture series [5] . The picture series technique allows educators to see the development of students' speaking skills. The picture series learning model positively impacts speaking ability [6]. There is a significant difference between classes that use conventional learning methods and classes that use the picture series learning method [7] . Vlog projects can improve speaking skills by an average of 27% . Vlog projects are the preferred learning method, and a student's level of desire for the method is 89%[8] . Like previous research, research [9] also revealed that vlog projects are an effective learning method to improve students' speaking skills. Some previous studies have discussed the effectiveness of picture series and vlog project learning methods to improve students' ability to speak English [6] . Few discuss which comparison is more effective than the picture series and vlog project. Research that compares picture series and vlog project learning methods is research [5] . The study's results [5] stated that the vlog project was more effective than the picture series. Comparing effective learning methods in speaking skills can be done with machine learning techniques [10] . Research [11] applied a machine learning algorithm, K-Means, to evaluate effective learning methods. K-Means effectively cluster and evaluate student performance [12] . This research continues the research conducted by previous research [5] . Since K-Means works effectively and few studies compare the learning methods of picture series and vlog projects, this research aims to group students based on pre-learning and post-learning results. This research has contributed in terms of verifying cluster changes that have not been carried out by previous studies. Grouping children is expected to determine the changes or consistency of childre

FAQs

What teaching techniques are compared to improve speaking skills in this study?

The study compares conventional learning, vlog project learning, and picture series learning for developing students’ speaking skills.

How does the study use machine learning to analyze speaking skill effectiveness?

K-Means clustering is applied to students’ pre-test and post-test scores, grouping learners based on score patterns and revealing changes in group membership.

What data is used to measure differences between student groups?

The research uses pre-test and post-test scores collected from three classes, then evaluates how clusters change after each teaching method.

Aug 3, 2026

Upload by Sage