Journal of Research Methodology

Undergraduate Students' Perspectives on Death Management Policy: A Secondary Data Analysis of Classroom Learning Traces

Titinan Pewnil — 2026-06-25

As Thailand transitions into an aged society with deaths exceeding births since 2021, death management policy has become an urgent public issue. Yet the perspectives of younger generations on this topic remain largely underrepresented in policy discourse. This article aims to: (1) examine the characteristics of learning processes and group work in designing death management policy proposals; (2) analyze the main issues and components of death management policy proposals from undergraduate students’ perspectives; and (3) analyze students’ perspectives toward death as reflected through their policy proposals. Using qualitative secondary data analysis (QSDA), we analyzed death management policy proposals produced by 20 student groups comprising 160 students in a Population and Development course during Semester 2 of Academic Year 2023. These proposals were generated through flipped classroom and collaborative learning activities. Data were analyzed using deductive-inductive content analysis and thematic analysis guided by the Policy Triangle Framework. The findings revealed three main points. First, student groups composed of members from multiple faculties and mixed-year levels proposed more diverse and comprehensive policies than homogeneous groups. Students strategically used humor in classroom presentations to reduce tension around death discussions. Second, students proposed policies covering three phases: before death, dying, and after death. These proposals encompassed all four dimensions of the Policy Triangle Framework, including context, content, process, and actors. Third, students’ perspectives on death reflected three transformations: acceptance of death as a public issue open for discussion, awareness of inequality in accessing dignified death, and a challenge to existing power structures in society that control death. This study demonstrates that qualitative secondary data analysis of learning activities is an appropriate approach for studying socially sensitive topics. This method enables access to authentic perspectives from young adults and may reduce social desirability bias, while also providing opportunities for the younger generation to participate in end-of-life policy development aligned with Thailand’s demographic transition context.

Designing and Validating an Evidence-Based Pedagogy Framework: A PRISMA–Delphi–IRR Approach for Technical Education

Arthur Dela Peña — 2026-06-25

Instructional quality in aviation maintenance training (AMT) is commonly regulated through content requirements, with limited attention to pedagogy and instructional assurance. This study develops and validates RAMP-ACE, an auditable framework for improving the reliability and transparency of competency-based teaching. A PRISMA-guided systematic review initially identified seven preliminary pedagogical domains. Through regulatory triangulation and Delphi refinement, overlapping standardization/calibration and quality-assurance constructs were consolidated, producing a final six-domain framework. The candidate item pool progressed from 14 indicators in v1 to 23 indicators in v2, with 21 retained after two indicators failed Delphi thresholds. Content validity was examined through a two-round asynchronous Delphi process (Round 1 N = 14; Round 2 N = 13), followed by a pilot audit assessing interrater reliability, rating time, artifact availability, and corrective-action closure. Delphi findings showed strong consensus, with most indicators meeting I-CVI ≥ 0.78, S-CVI/Ave = 0.89, and S-CVI/UA = 0.35. Pilot testing produced κ/ICC values of 0.72–0.86, median dossier rating time of 38 minutes, and artifact availability above 85% for key records. Corrective actions were mostly closed within two weeks, supporting operational feasibility and transferability to technical and competency-based education contexts.

Development of Computer Software for Criterion-referenced Item Analysis

Nhabhat Chaimongkol — 2026-06-25

This research aimed to develop and assess the quality of computer software for criterion-referenced item analysis, designed to align with Thailand’s standards-based basic education curriculum.
A research and development (R&D) methodology was employed, comprising two phases: Phase 1, software development, and Phase 2, software quality assessment. The research instruments included a heuristic evaluation checklist for assessing the appropriateness of the software interface and a user experience questionnaire. The participants who tested the software consisted of 50 users. The results revealed the following: 1) The developed software could serve as a replacement for existing software whose analytical outputs may be difficult to interpret for users without a background in educational measurement and evaluation, while also being compatible with current operating systems. The software developed in this study, named i-ACT (Item Analysis for Criterion-Referenced Test), was developed using the Visual Basic programming language and can be used without installation (portable software). The key item analysis results generated by the software include the difficulty index, discrimination index, and reliability coefficient, based on criterion-referenced test analysis concepts. 2) Regarding software quality assessment, users reported a positive experience with the software. Users rated the software highly in terms of performance efficiency (M = 6.26, SD = 0.89) and ease of learning (M = 6.24, SD = 0.71). In addition, users perceived the developed software as appealing (M = 6.20, SD = 0.81).

Object Detection Using Computer Vision

Jiramate Rujikornhirun — 2026-06-25

Object detection has become an important technology in artificial intelligence and computer vision. Many industries now use this technology, including self-driving cars, medical image analysis, security systems, and manufacturing. This paper presents a detailed study of object detection methods through three main parts. The first part covers the basic theories and concepts of object detection. We explain the differences between image classification and object detection, and describe the main parts of detection systems. We also discuss how to measure performance using average precision (AP) and mean average precision (mAP). The section includes a review of how detection models have developed over time, from older methods to modern one-stage and two-stage approaches. We provide practical examples using Faster R-CNN. The second part focuses on the YOLO (You Only Look Once) algorithm, which represents single-stage detection methods. YOLO is popular because it works fast while keeping good accuracy. We explain how YOLO works, including how it divides images into grids and predicts bounding boxes and object types at the same time. We show detailed code examples for detecting objects in still images using YOLOv5. The last part extends to video object detection, which brings new challenges for real-time processing. We discuss various techniques to improve performance, such as model quantization, pruning, knowledge distillation, and combining object tracking with detection using DeepSORT. We demonstrate practical applications by showing how to detect and track cars in traffic videos. The demonstrations illustrate how pretrained object detection models, particularly Faster R-CNN and YOLOv5, can be applied to still images and video frames for instructional purposes. The article does not aim to provide a systematic benchmark comparison, but rather to present practical implementation examples for readers who are beginning to use object detection techniques.

Front Matter

Journal of Research Methodology — 2026-06-25