As the significance of data-driven decision-making is increasingly emphasized, statistical modeling has emerged as a key area in statistics education. While international studies have extensively elucidated how elementary students develop statistical ...
As the significance of data-driven decision-making is increasingly emphasized, statistical modeling has emerged as a key area in statistics education. While international studies have extensively elucidated how elementary students develop statistical reasoning, domestic studies lacks in-depth exploration of the micro-cognitive processes through which students develop their reasoning while performing modeling tasks. Given that primary students often employ informal and intuitive methods, it is necessary to explore the dynamic and intuitive reasoning processes manifested in their statistical modeling. Furthermore, considering the limited opportunities within the national primary curriculum to explore various facets of distribution, this study aims to interpret students’ reasoning regarding distribution.
The purpose of this study is to provide a profound understanding of how sixth-grade students develop their reasoning on distribution within the process of informal statistical modeling. To this end, the study tracks the development of conjecture models and data models constructed by student and anlyzes how the interaction between these two models facilitates informal statistical inference. By focusing on variability and distribution, the study specifically examines the emergence of aggregate reasoning, wherein students perceive data not as case-value but as an aggregates.
In terms of methodology, a qualitative case study was conducted with students in grades 4-6 from an elementary school in the subarb area, utilizing TinkerPlots for statistical instruction. The analysis focused on three sixth-grade students, employing the microgenetic method to examine diverse data sources. This study applies the Reasoning with Informal Statistical Models and modeling(RISM) framework proposed by Dvir & Ben-Zvi (2018), alongside the perspectives on data by Konold et al. (2015) to analyze shifts in students’ reasoning about distribution.
The findings are as follows: First, a dynamic interaction between students’ conjecture models and data models was observed. In the initial stages of inquiry, students' strong prior beliefs occasionally distorted data interpretation or led to a regression toward existing perspectives. However, as the inquiry progressed, students critically refined their initial conjecture models based on empirical evidence and actively structured new models, that aligned with the characteristics of the data. Second, this model interaction functioned as a core mechanism for deepening students' reasoning about distribution. Students' reasoning evolved from a classifier to an aggregate perspective. Ultimately, students recognized the provided data as a sample, critically evaluated its reliability and representativeness, and reached informal statistical inference by selecting evidence congruent with the context.
This study identifies that the interaction between conjecture and data models is a pivotal mechanism that deepens the understanding of distribution and drives informal statistical inference. These findings suggest the necessity of designing tasks that intentionally induce discrepancies between models and recognizing students' initial conjectures as essential pedagogical resources. Furthermore, teachers should foster a classroom culture focused on constructing evidence-based arguments rather than merely seeking correct answers, thereby supporting students in becoming rational inferrers who can refine their beliefs based on data.