International Journal of Institutional Research and Management

Development of an Automated Institutional Research System for Institutional Decision-Making in Kyushu Institute of Technology

Tetsuya Oishi — Sun, 10 Mar 2024 00:00:00 +0000

A system has been established at the Kyushu Institute of Technology (Kyutech) that enables both students and teachers to reflect on their learning history and complete self-evaluations using e-portfolio systems. However, the university has not yet implemented institutional research (IR) for decision-making at the institutional level. To address this gap, the Learning Teaching Center at the university decided to establish an Educational IR Support Group in FY 2022, aiming to strengthen IR in education. In addition to e-portfolio systems, various data are accumulated across different systems within the university. However, there were no integrated systems to utilize these data comprehensively, as various departments managed them individually. To overcome this, we introduced an extract transform load (ETL) tool as part of the initial IR system. Simultaneously, we implemented a data lake system as a centralized repository for storing data. The Educational IR Support Group is leading this project. This study begins by providing an overview of IR and how it should be conducted. It then presents the IR system being promoted at our university, including the introduction of the ETL tool and the establishment of a data lake. Finally, we describe our future plans and developments, which involve the potential introduction of a data warehouse and business intelligence tools.

A Case Study of Using the Structural Topic Model to Grasp Actual Learning through Free-Writing Reports

Tatsuya Tsumagari, Yoko Nakazato, Takashi Tsumagari — Mon, 12 Aug 2024 07:59:33 +0000

This paper reports the results of an analysis using structural topic model (STM) of career awareness to examine learning in first-year career education courses. The subject of this study was a first-year career education course offered by a required course at public University A in Japan. The free-writing reports of 1,780 first-year students were included in the analysis.

We examined students' career awareness from two perspectives using STM. First, we examined stable career awareness that does not fluctuate from year to year. We identified whether stable career awareness was created by the same syllabus through changes in external factors. Second, we examined the impact of differences in student types on learning.

Fifteen topics were extracted as career awareness. Among those topics, six topics were found to be stable career awareness topics regardless of year. In addition, when students were classified into five types according to their interest in the lectures, eight topics were found to apply to constant career awareness, regardless of their interest. The three topics were independent of both year and student type.

The analysis of students' free-writing reports using STM is useful as an assessment method for grasping the reality of student learning.

A Study of Enrollment Projections for USA Higher Educa-tion Institutions

Emma Gyasi, Felix Famoye, Carl Lee, Robert Roe — Tue, 06 Feb 2024 16:08:52 +0000

This study provides results from a survey on enrollment projections, methods, metrics, timing, and model among public 2-year and 4-year higher education institutions in the United States. The data are from 127 public, 4-year and 73 public, 2-year institutions surveyed in spring and summer 2021. The results are summarized on various aspects of the process for developing enrollment projection numbers from the factors considered, the type of enrollment models used, methods and modeling techniques implemented, and the involvement of campus offices. These findings will help provide details on current enrollment models, methods and modeling techniques implemented, and campus offices' involvement in enrollment projections in higher education institutions. The study reveals, there is no vast difference in how public, 4-year and public, 2-year institutions oversee enrollment projections. Almost all institutions build and develop their enrollment models in-house. The most widely used software for modeling and presenting enrollment projections is Microsoft (MS) Excel. The top three modeling techniques implemented in enrollment projection are Time series models, Markov chain models, and Linear regression models. Multiple offices in the institutions participate in the process of producing enrollment projection numbers.