This paper introduces an educational tool titled Mixed Methods Research for Nursing Researchers, which was developed to address the challenges associated with learning and implementing mixed methods research. Based on the experiences of nursing researchers in Japan and abroad, this paper examines the background, objectives, structure, and significance of the tool. This e-learning program was developed primarily by board members of the Japan Society for Mixed Methods Research (JSMMR) and was designed as an educational tool for systematically learning mixed methods research. It incorporates the traditional Japanese learning model of "Shu-Ha-Ri," allowing learners to progress through structured, step-by-step learning based on their proficiency levels. At the 10th Annual JSMMR Conference Open Forum in 2024, representatives of the program developers and test users participated in discussions, sharing their experiences with mixed methods research while evaluating the usability of the program. This paper reports on the forum’s discussions and examines the possible impact of this e-learning program on the learning process of mixed methods research. Through this program, we aim to enable learners to actively develop research designs suited to their research objectives and effectively apply mixed methods research in practice.

This study proposes a novel mixed method research approach that integrates the epistemology and methodology of Bayesian statistics (including application in AI). To reframe each research paradigm: quantitative research (frequency theory) is positioned as hypothesis testing by deduction (theory driven); qualitative research as hypothesis generating via abduction and self-validation through qualitative induction (data interactive); and Bayesian statistics as hypothesis selection through quantitative induction (data driven). The proposed mixed method approach involves the following steps: 1) Designing the overall research framework and building the database using qualitative research; 2) Applying Bayesian statistics to the database to identify the most plausible model and narrow down key hypotheses and variables; 3) Testing these key hypotheses through quantitative research and summarizing the overall conclusions; 4) Feeding significant findings back into the research project to iteratively refine and restart the process.

A research project funded by Kaken, the Grant-in-Aid for Scientific Research, and conducted by board directors from the Japan Society for Mixed Methods Research (JSMMR), has successfully developed a comprehensive e-learning program for Mixed Methods Research (MMR) focusing on nursing researchers. As a result, we have publicly begun offering the program in spring 2025. In this academic journal published by JSMMR, we have launched the series, “Method Cafe: Mixed Methods Dialogues in Relay,” to provide materials for learning MMR that support the e-learning content further, such as providing opportunities to explain research designs.

As the first in the series, this paper aims to deepen readers' understanding by introducing two mixed-method research studies that also applied the positive deviance (PD) approach and examining and explaining them from the perspectives of the PD approach and MMR.

Objective: To survey how dizziness patients choose specialty clinics and what the final diagnoses are in a clinical training hospital.

Methods: We conducted the study in the designated secondary emergency hospital in the urban area. We retrospectively surveyed patients who visited three of the departments: Internal Medicine and General Medicine (IMGM), Emergency Medicine (EM), and Otolaryngology (ENT), with the chief complaints of dizziness between February and July 2022. The inclusion criteria were patients with medical records of either two dizziness-related keywords, ‘Memai’ in hiragana or kanji. We analyzed quantitative data using correspondence analysis and frequency analysis. We used thematic analysis for the qualitative data from the semi-constructive interview with the head nurse in charge of outpatient clinic management. We used the convergent mixed methods design to draw metainferences.

Results: In a half year, IMGM doctors saw 217 patients, particularly with stable symptoms or variable systemic illnesses, including non-emergent disease or otolaryngologic diseases. Forty percent of patients whom ENT doctors saw had otolaryngologic diseases with ear symptoms. EM doctors saw emergent patients with severe symptoms requiring hospitalization or vital sign abnormalities. Moreover, twenty percent of the patients had benign paroxysmal positional vertigo. All clinics used ruling-out diagnoses.

Conclusions: We clarified the clinical reality of how dizziness patients select specialty clinics, including their numbers and final diagnoses.