The Science of Design: Introducing Design of Experiments (DoE) through a Hydrogel-Based High School Chemistry Investigation

Résumé fourni par la source

Abstract Creating lessons that connect classroom content to real-world application can enhance high school students’ interest in STEM curricula. The scientific method is one such component of the curriculum that can be meaningfully connected to techniques and practices used across a range of STEM fields. Specifically, Design of Experiments (DoE) is a statistical process widely used in STEM industries and research and development contexts to efficiently collect and analyze scientific data, yet it is rarely taught at the high school level. Importantly, DoE also introduces foundational concepts in structured data generation, multivariable analysis, and pattern recognition that underpin emerging artificial intelligence (AI) and machine learning (ML) approaches. To bridge this gap between methodological understanding and real-world application, this study implemented a hands-on DoE-based lesson focused on polymer hydrogels, which was assessed in a high school science classroom with students across grades 9–12. Pre- and postlesson assessments, consisting of multiple-choice, short-answer, and conceptual understanding questions, were used to evaluate shifts in students’ content knowledge and understanding of the DoE approach. The median correct response rate for the conceptual understanding section increased from 33.33% to 66.67% between the pre- and postassessments. Additionally, while response rates for conceptual questions increased following the intervention, a statistically significant improvement in response accuracy was not observed. This pilot study serves as a valuable starting point for designing STEM lessons that introduce students to DoE as a framework for experimental design, while also providing insight into student understanding of this approach─an essential skill set for preparing students for future STEM workforce demands.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.