Embedded Files
DISSERTATION
DISSERTATION
Automatic Item Generation Using Multi-Stage Agentic Systems
Ph.D. Dissertation, University of North Carolina at Greensboro (In Progress)
Focuses on leveraging large language models (LLMs) and agentic workflows to generate high-quality exam items.
Going from template-based AIG to non-template-based AIG
Multiple Agents (Cognitive Model Agent, Item Generation Agent, and Reviewer Agent)
Integrates psychometric validation frameworks (Responsible use of AI)
Built SME Review System
Field-tested the AI Items for psychometric characteristics
Tools (Python, VS Code, LangChain, LangGraph, OpenAI Enterprise)
Aims to advance responsible AI in educational measurement by blending statistical rigor with innovative AI methods.
INTERNSHIP RESEARCH
INTERNSHIP RESEARCH
Physician Assistant Education Association (PAEA)
Physician Assistant Education Association (PAEA)
Contributed to psychometric analyses, standard setting workshop, and automated test assembly.
Built R Shiny applications for grade conversion and operational assessment support.
National Commission on Certification of Physician Assistants (NCCPA)
National Commission on Certification of Physician Assistants (NCCPA)
Designed an AI-powered medical item generation system for the PANRE-LA exam.
Developed a RAG system (integrating PubMed abstracts and textbooks).
Presented results to the Board of Directors, showing AI-generated items comparable to SME-written ones.
FocalPoint Education
FocalPoint Education
Conducted a comparative study of eight LLMs for automated essay scoring.
Evaluated psychometric properties using Quadratic Weighted Kappa (QWK), ICC, and RMSE.
Contributed to a special issue manuscript on AI and writing assessment.
Publications
Peer-Reviewed:
Makinde, H. S., Makinde, A. I., Usman, M. A., Adegoke, H., Makinde-Isola, B. A., Lawal, W., & Jimoh, I. T. (2025). The readability paradox: Can we trust decisions on AI detectors? Technium Education and Humanities, 11, 181-195. https://doi.org/10.47577/teh.v11i.12946
Under Review:
Murphy, S., Lynch, M., D'Brot, M., & Makinde, H. (In-Press). Evaluating the accuracy, reliability, and applicability of multiple large language models in automated scoring for writing assessments. Assessing Writing. Manuscript ASW-D-25-00432.
Makinde, H., Adegoke, H., & Mojoyinola, M. (In-Press). PsychMet- measurement foundational competencies ChatBot. In Proceedings of the NCME AIME Conference 2025. National Council on Measurement in Education.
Feyijimi, T., Oyeniran, D., Apata, O., Makinde, H., Adegoke, H., Ajamobe, J., & Dadzie, J. (In-Press). Augmented measurement framework for dynamic validity and reciprocal human-AI collaboration in assessment. In Proceedings of the NCME AIME Conference 2025. National Council on Measurement in Education.
KEY PROJECTS
PsychMet ChatBot
Co-created a chatbot to support psychometric foundational competencies.
Accepted for presentation at NCME AIME Conference (2025).
Doctor AI Academy: 30 DAYS 30 TERMINOLOGIES
For about 30 days, I successfully demystified 30 essential AI concepts, from foundational Artificial Intelligence to cutting-edge technologies like Large Language Models (LLMs) and Generative AI. Each day featured a new term explained in simple, accessible language with real-world examples, making complex AI concepts understandable for everyone. This comprehensive journey created a complete AI vocabulary guide, empowering professionals, students, and curious minds to confidently navigate the AI landscape.
#30Days30Terms #AIEducation #DoctorAIAcademy #LLM #GenerativeAI #MachineLearning
DoctorAI's 5-Week AI_101 Training: From Zero to AI Hero! 🚀
I'm excited to launch this comprehensive 5-week training program that transforms complete beginners into confident AI practitioners. Starting with Python & R fundamentals in Week 1, we build a solid programming foundation before diving into the mathematical backbone of AI with Statistics and Math in Week 2. Week 3 introduces Machine Learning basics, where participants learn to build their first predictive models, followed by Week 4's exploration of Deep Learning and Neural Networks - the technology powering today's AI revolution. The journey culminates in Week 5 with Generative AI and Large Language Models, where students learn to work with cutting-edge technologies like ChatGPT and DALL-E.
ALDDN Project Success: From Field Implementation to Strategic Data Leadership (2020-2025)
My journey with the Advancing Local Dairy Development in Nigeria (ALDDN) project represents a comprehensive evolution from on-ground implementation to high-level data strategy consulting. Beginning in 2020 as a field-based data collection specialist in Nigeria, I managed the entire Survey Solutions platform deployment, training over 100 enumerators across Northern Nigerian states and ensuring real-time data quality through innovative validation systems. As Survey Solutions Manager and Data Processing lead from 2020 with Ipsos Nigeria, I elevated the project's impact by implementing advanced data quality assurance frameworks, coordinating with Mathematica Research for endline evaluations, and aligning multi-phase datasets (baseline, midline, endline) to demonstrate the program's transformative impact on smallholder dairy farmers. This 5-year commitment resulted in robust evidence that shaped dairy sector policies, improved livelihoods for thousands of farmers, and established scalable models for agricultural development across sub-Saharan Africa.
#ALDDN #DataManagement #SurveySolutions #ImpactEvaluation #AgriculturalDevelopment #Nigeria #IpsosNigeria#Mathematica
Page updated
Google Sites
Report abuse