Instructor-Personalized WolframAlpha-Supported Instruction for Learning Mathematical Functions: A Longitudinal Quasi-Experimental Learning Analytics Study
Clodoaldo Ramos Pando, Teófilo Félix Valentín Melgarejo, Flaviano Armando Zenteno Ruiz, Armando Isaías Carhuachín Marcelo, Raúl Malpartida Lovatón, Víctor Luis Albornoz Dávila, Ulises Espinoza Apolinario, Rogelio Amancio Landaveri Martínez, Pablo Lolo Valentín Melgarejo, Wilmer Napoleón Guevara Vásquez, Nely Teresa Aldana Taniguche, Pablo Lenin La Madrid Vivar, Liz Ketty Bernaldo Faustino, José Rovino Alvarez Lopez
Source abstract
Computational knowledge tools are increasingly integrated into higher education, but their educational value may depend less on the technology itself than on the pedagogical structure surrounding its use. This longitudinal study compared four instructional conditions for undergraduate learning of mathematical functions: traditional instruction (TI), unstructured WolframAlpha use (UWA), scaffolded WolframAlpha-supported instruction (SWA), and instructor-personalized WolframAlpha-supported instruction (IPWA). WolframAlpha was treated as a non-adaptive computational knowledge engine rather than as an autonomous adaptive tutoring system. In IPWA, personalization was implemented by instructors through predefined pedagogical responses to observed learner indicators and was not generated autonomously by WolframAlpha. A longitudinal four-arm quasi-experimental study was conducted with 600 first-semester undergraduate students at the National Daniel Alcides Carrión University, Peru. The eight-week intervention comprised traditional instruction, unstructured WolframAlpha use, scaffolded WolframAlpha-supported instruction, and instructor-personalized WolframAlpha-supported instruction, followed by retention and transfer assessments at Weeks 16 and 20. Learning outcomes were assessed across five occasions, and behavioral records were collected during WolframAlpha-supported sessions. The instructor-personalized condition showed the highest adjusted achievement trajectories and stronger maintenance of learning at the follow-up assessments. Self-regulated learning, reflection frequency, prompt complexity, and prior achievement were positively associated with mathematics achievement, whereas mathematics anxiety was negatively associated. Platform-interaction analyses identified transitions toward more exploratory and verification-oriented patterns; however, these records did not capture learning activities undertaken outside WolframAlpha. The findings suggest that instructor-mediated personalization surrounding a computational tool may support mathematics learning, although the quasi-experimental design does not permit definitive causal conclusions.
Evidence graph
No public relationships recorded yet.
Integrity note: This page is a factual metadata record created by deterministic ingestion. It is not a claim that the work moves a mathematical frontier or has been independently verified.