IIAI Letters on Institutional Research

The Dialogic Dual-Instructor Model (DDIM)

Kunihiko Takamatsu — 2026-01-31

This study investigates which instructional formats most effectively promote student understanding and engagement in post-COVID-19 higher education. While the pandemic accelerated diversification in teaching modes including synchronous online classes, on-demand videos, and blended formats, limited empirical evidence exists comparing their effectiveness. Drawing on dialogic pedagogy, STEAM education principles, and the eduinformatics framework, we examine monologic single-instructor lectures, teaching assistant-supported classes, and dialogic co-teaching formats. We introduce the Dialogic Dual-Instructor Model (DDIM), a collaborative teaching approach implemented across statistics courses at Kobe Tokiwa University since 2017. DDIM involves two instructors engaging in structured dialogue during instruction, with one primarily presenting content while the other poses questions, requests clarifications, and offers alternative perspectives that mirror student thinking processes. This approach has been successfully adapted across face-to-face, audio-only, and on-demand video formats. Based on qualitative analyses of classroom implementations and instructor reflections, our analysis, synthesizing prior research on tutorial-style videos and dialogic practices in STEAM contexts with our collaborative statistics education practice, suggests that DDIM represents an effective instructional format particularly for conceptually demanding university courses, for fostering student engagement and understanding in contemporary university education.

Effects of Career Change Opportunities for Graduates After Graduation from an Educational Institute

Yuya Yokoyama — 2026-01-31

In modern society, where drastic transformations of social structures are occurring, relearning new knowledge and skills for adults is becoming progressively vital thanks to the widespread proliferation of various educational methods. Under these circumstances, our study aims to analyze the underlying factors of study motivation of adult learners. To reveal the factors and relationships between before enrollment and after graduation, we performed multivariate analysis on a questionnaire targeted at general adult learners. Based on these observations, we created a questionnaire aimed at the graduate students of our affiliation. Factor analysis has yielded five factors reflecting the curriculum of our institute. Besides the questions analyzed in our previous work, we also posed a question about usefulness for work, which could be an important clue to evaluating the effectiveness of our curriculum. Therefore, in this paper we performed cluster analysis to examine the tendency in the graduates’ career changes after graduation. In executing cluster analysis, three sets of feature values were analyzed with four clusters. As a result, it can be suggested that among clusters with high competence, the entire opposite groups can be observed; that is, some graduates stay in the same organizations as before, while others seek new environments.

Evaluating the Effects of Pre-University Education Using Propensity Score Matching

Keita Nishiyama — 2026-01-31

This study uses propensity score matching to explore the causal relationship between pre-university education and first-year GPA. Given the challenges of conducting randomized controlled trials in educational settings due to ethical concerns and practical limitations, observational studies often become necessary. Propensity scores, initially proposed by Rosenbaum and Rubin, enable a more reliable estimation of causal effects by simulating an experimental framework in observational data. This method adjusts for covariate distributions between treatment and control groups, allowing for statistically comparable conditions. The findings support that pre-university education significantly improves first-year GPA, demonstrating the method's effectiveness in educational research and highlighting its potential for broader application across various academic disciplines.

Beyond Bilateralism and Multilateralism?

Patrick Shorb — 2026-01-31

Be it through the establishment of bilateral partnership institutions or multilateral quality enhancement projects, over the past two decades, the Japanese government has invested tens of millions of dollars in official development assistance (ODA) to improve Southeast Asian university-level engineering and technology education. Drawing upon ODA project-related documents and site visit interviews, this study explores why Japan devoted so much ODA resources to such a narrow goal. On the one hand, these projects continued Japan’s long-standing commitment to the hitozukuri (“human resource development”) of its Southeast Asian partners. On the other hand, they represent an important departure from past approaches, as this aid, through its increasing emphasis on goal-oriented multilateral collaborations, has also garnered widespread regional goodwill at relatively low cost.

High School Student Sessions at Academic Conferences in Japan: Status and Participant Attributes

Noa Iwai — 2026-01-31

In the Japanese education system, the 2018 revision of the High School Curriculum Guidelines emphasized inquiry-based cross-disciplinary study, encouraging students to engage in substantive research and presentation activities beyond the classroom. This study aims to clarify the current status of high school student sessions at academic conferences in Japan and to investigate the attributes of the presenting schools. A two-part survey was conducted: (1) Analysis of websites of leading Japanese academic societies to identify those holding high school student sessions; (2) Examination of programs and affiliated schools of presenters in these sessions, linking them with official school data to analyze school types, locations, and participation patterns. Results showed that high school sessions mainly exist in natural science fields, with participating high schools including both public and private institutions, and the geographic spread expanding beyond metropolitan Tokyo. Many schools participate only once, while about 20% engage repeatedly, indicating varied continuity. This study offers novel data on Japanese high school students’ academic presentations, an area rarely examined in Educational Data Science or Library and Information Science. It contributes to clarifying high school students' academic communication practices. It provides foundational insights for academic societies and universities hosting academic conferences to consider whether to open their events to high school students.

Outcomes of Interdisciplinary Graduate Education

Ming Li — 2026-01-31

Interdisciplinary graduate education has become increasingly important in addressing complex global challenges that transcend traditional disciplinary boundaries. This study examines the outcomes of such education programs through the lens of Tomlinson’s Graduate Capital Model, which encompasses human, social, cultural, identity, and psychological capitals. Focusing on a case study at a Japanese national university, qualitative data were collected through semi-structured interviews and free-form questionnaire with program graduates. Using KH Coder for quantitative text analysis, the results revealed that interdisciplinary education strengthens graduates’ academic, social, and personal development. It broadens knowledge and skills by combining ideas and methods from different fields, improving research ability, employability, and adaptability. Collaborative learning across disciplines helps students build professional networks and appreciate diverse academic and workplace cultures. These experiences also shape their professional identity, foster career flexibility, and build confidence and resilience. Overall, interdisciplinary learning fosters well-rounded development by cultivating both expertise and transferable skills essential for lifelong learning and success.

Design and Implementation of a Cyclic Dropout Preven-tion Model Using Institutional Research Data

Naruhiko Shiratori — 2026-01-31

This position paper proposes a cyclic model for dropout prevention in higher education, integrating data-driven prediction and student support practices. The model connects six key phases—data consolidation, time-series dropout risk prediction, student status monitoring, classification of student trajectories, targeted intervention, and evaluation of support effectiveness—into a continuous improvement cycle. Grounded in institutional research (IR), the model utilizes attendance records, academic performance, and pre-admission data to estimate dropout probabilities and classify students using clustering techniques such as X-means. Based on these classifications, tailored interventions including early alert systems and enhanced first-year education programs are implemented. The effectiveness of these interventions is evaluated through changes in attendance and academic outcomes, enabling feedback into the model for refinement. This framework aims to bridge the gap between predictive analytics and practical student support, offering a scalable and adaptable approach for universities seeking to reduce dropout rates and improve student success.

A Time-Constrained Analysis of Dynamic Early Warning Systems for Academic Risk Prediction

Shintaro Tajiri — 2026-01-31

Implementing effective academic support for mandatory first-year courses requires precise decision-making about when to intervene, with whom, and with what level of certainty. This study extends our previous static prediction model (AUC=0.878 [1] using enrollment data alone) by addressing its key limitation: the inability to answer operational questions about intervention timing. Using data from a mandatory Information Literacy course at Hokuriku University (N=335, Economics and Management faculty, 2022-2023), we developed machine learning models that incrementally add dynamic formative assessment data from weeks 2-8 to static enrollment information. Under strict time-constraints preventing data leakage, we evaluated models using Recall@Precision≥0.90—a practical metric balancing intervention resource constraints with student rescue effectiveness. Results demonstrate that minimal behavioral features from weeks 2-8 (submission rates, task completion counts) significantly improve Recall@P≥0.90 from 1.6% to 3.2%, doubling rescue capacity while providing weeks of intervention lead time.