Hi everyone! My name is Oluwaseun Farotimi, and I am a Ph.D. candidate in the Methodology, Measurement, and Analysis program at the University of Central Florida. My research revolves around mathematical modeling, machine learning, and meta-analysis.
My curiosity about machine learning began with my quest to harness data storytelling more efficiently. Every piece of data truly tells a story! Effective data storytelling will continue to guarantee an adept understanding of underlying factors and effects of interventions and policies. For example, survey data about student achievement are collected to uncover critical insights into factors that influence academic performance, grades, student engagement, etc. Hence, it is essential to convey precise, actionable data insights through storytelling! The ability to think rigorously, implement complex algorithms, and convey precise, actionable insights through storytelling is essential.
Lessons Learned
Effective storytelling with machine learning requires some skill
When it comes to utilizing machine learning models for data analysis, there are crucial aspects to consider. Data analysis is a cornerstone of educational research, driving informed decision-making processes that enhance learning outcomes and institutional effectiveness. A data-driven approach to teaching and learning enables faculty, schools, and universities to develop personalized learning pathways for students with diverse academic backgrounds and needs. Therefore, it is imperative to acquire technical machine-learning expertise.
Key points to note for data storytelling
The following are key points to note: understand your audience (technical vs. non-technical), embrace data visualization, use a compelling storytelling pattern to elucidate findings, interpret model performance results for your audience type, contextualize the data with a narrative, and provide strategic data-driven recommendations.
Hot Tips
Be assertive with your data storytelling
Numbers alone don’t tell the stories! Quantitative metrics such as accuracy, precision, f1 score, and recall are measures used to evaluate how well machine learning models perform in any predictive, classification, or regression task. However, it’s the researchers’ excellent narrative skills that truly assert the data story.
Each machine-learning model has varying mechanisms
Fundamentally, there are two main types of machine learning: supervised and unsupervised. The key difference is that supervised learning uses labeled data, while unsupervised learning does not. Common machine-learning models are linear regression, decision trees, random forests, support vector machines, and k-nearest neighbors; each model has a different underlying mechanism/algorithm and performs differently given the dataset.
The American Evaluation Association is hosting Quantitative Methods: Theory and Research Evaluation Week. The contributions all this week to AEA365 come from our Quant TIG members. Do you have questions, concerns, kudos, or content to extend this AEA365 contribution? Please add them in the comments section for this post on the AEA365 webpage so that we may enrich our community of practice. Would you like to submit an AEA365 Tip? Please send a note of interest to AEA365@eval.org. AEA365 is sponsored by the American Evaluation Association and provides a Tip-a-Day by and for evaluators. The views and opinions expressed on the AEA365 blog are solely those of the original authors and other contributors. These views and opinions do not necessarily represent those of the American Evaluation Association, and/or any/all contributors to this site.