Explorando rasgos de la producción oral de estudiantes de Nt1 y Nt2, a través de una herramienta que utiliza inteligencia artificial. Implicancias teóricas y aplicadas de los hallazgos para el aprendizaje y la metacognición.
Loading...
Date
2025
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de Concepción
Abstract
La presente investigación tiene como propósito analizar la producción oral infantil en los niveles NT1 y NT2 desde las dimensiones lingüística, emocional y metacognitiva, empleando herramientas de inteligencia artificial integradas en el MetaSistema. El objetivo fue identificar diferencias entre niveles educativos y condiciones del neurodesarrollo (neurotípica y neurodivergente), además de evaluar el aporte de la IA para una medición objetiva del discurso.
Se utilizó un diseño cuantitativo, exploratorio y comparativo. Las producciones orales fueron procesadas mediante TRUNAJOD, COMUNICA, INTERPRETA y ANALIZA, obteniendo indicadores relacionados con coherencia, densidad léxica, prosodia, emociones y metacognición. Los análisis aplicados incluyeron pruebas t de Student, análisis discriminante, ANOVA bifactorial, correlaciones y percentiles.
Los resultados muestran que NT2 presenta un desempeño más consolidado en las tres dimensiones: mayor coherencia y organización sintáctica en lo lingüístico, una prosodia más regulada en lo emocional y niveles superiores de verbalización metacognitiva. En NT1 predominan producciones más simples, expresivas y menos reguladas. Respecto a las condiciones del neurodesarrollo, los niños neurotípicos exhibieron mayor diversidad léxica, estabilidad prosódica y puntajes metacognitivos más altos, mientras que los neurodivergentes mostraron mayor variabilidad interna y diferencias emocionales específicas. Integralmente, la IA se posiciona como una herramienta válida para evaluar objetivamente la producción oral infantil.
The present study aims to analyze children’s oral production in NT1 and NT2 levels across linguistic, emotional, and metacognitive dimensions, using artificial intelligence tools integrated into the MetaSystem. The objective was to identify differences between educational levels and neurodevelopmental conditions (neurotypical and neurodivergent), as well as to evaluate the contribution of AI to an objective assessment of oral discourse. A quantitative, exploratory, and comparative design was employed. Children’s oral productions were processed through TRUNAJOD, COMUNICA, INTERPRETA, and ANALIZA, obtaining indicators related to coherence, lexical density, prosody, emotions, and metacognition. The analyses included Student’s t-test, discriminant analysis, two-way ANOVA, correlations, and percentile calculations. The findings indicate that NT2 demonstrates more consolidated performance across all three dimensions: greater coherence and syntactic organization in the linguistic domain, more regulated prosody in the emotional domain, and higher levels of metacognitive verbalization. In contrast, NT1 productions tend to be simpler, more expressive, and less regulated. Regarding neurodevelopmental conditions, neurotypical children showed greater lexical diversity, prosodic stability, and higher metacognitive scores, while neurodivergent children exhibited greater internal variability and specific emotional differences. Overall, AI proved to be a valid tool for objectively and multidimensionally assessing children’s oral production.
The present study aims to analyze children’s oral production in NT1 and NT2 levels across linguistic, emotional, and metacognitive dimensions, using artificial intelligence tools integrated into the MetaSystem. The objective was to identify differences between educational levels and neurodevelopmental conditions (neurotypical and neurodivergent), as well as to evaluate the contribution of AI to an objective assessment of oral discourse. A quantitative, exploratory, and comparative design was employed. Children’s oral productions were processed through TRUNAJOD, COMUNICA, INTERPRETA, and ANALIZA, obtaining indicators related to coherence, lexical density, prosody, emotions, and metacognition. The analyses included Student’s t-test, discriminant analysis, two-way ANOVA, correlations, and percentile calculations. The findings indicate that NT2 demonstrates more consolidated performance across all three dimensions: greater coherence and syntactic organization in the linguistic domain, more regulated prosody in the emotional domain, and higher levels of metacognitive verbalization. In contrast, NT1 productions tend to be simpler, more expressive, and less regulated. Regarding neurodevelopmental conditions, neurotypical children showed greater lexical diversity, prosodic stability, and higher metacognitive scores, while neurodivergent children exhibited greater internal variability and specific emotional differences. Overall, AI proved to be a valid tool for objectively and multidimensionally assessing children’s oral production.
Description
Tesis presentada para optar al grado de Educador/a de Párvulos.
Keywords
Análisis lingüístico, Metacognición, Niños, Inteligencia artificial