Impact of computer modeling on learning and teaching systems thinking

Abstract

Researchers have found that computer modeling fosters the learning of causal mechanisms in systems, an important crosscutting concept in science that many novice learners find challenging. Despite the research that highlights the role of teacher's instructional practices in enacting computer tools, few studies have considered teachers' use of computer modeling and its implications for student learning in classroom interactions, compared to interactions without computer tools. In this study, we examine (a) the impact of computer modeling on students' understanding of causal links in decomposition and (b) classroom interactions with use of computer modeling. We employed a quasi‐experimental design with eight middle school science classes that served predominately Latinx students. The random treatment was at the class level (computer modeling; n = 60, four classes) and control (paper modeling; n = 59, four classes). Analyses incorporated student preassessment and postassessment, classroom observations, and audio‐recorded modeling instruction. Results indicate that compared to paper modeling, computer modeling enriched systems thinking, particularly students' ability to provide causally coherent statements in explaining scientific ideas and evidence. Enactment of computer modeling may be associated with a shift in classroom interactions to include more invitation for students' elaboration of causal systems. We discuss aspects of computer modeling that may foster systems thinking, with implications for the future design of tools and curricula.

Building a computational model of food webs: Impacts on middle school students' computational and systems thinking skills

“…Although findings from learning progression areas provide a clear distinction on how students learn food webs and progress throughout their educational years, many researchers studied how appropriate instructional strategies could aid students in seeing more sophisticated relationships in an ecosystem (e.g., Grotzer & Bell, 2003; Lehrer & Schauble, 2012; Nguyen & Santagata, 2021). Notably, Sumarni et al (2017) compared the business‐as‐usual condition to a paper‐based drawing modeling condition and examined the impacts on students' argumentation skills within the food webs unit.…”

“…The authors found out that their participants developed a more advanced conceptual understanding of interdependence in an ecosystem after a 2‐week‐long intervention. More recently, Nguyen and Santagata (2021) conducted a quasi‐experimental study of sixth graders learning about decomposition in a food web. The authors found that students had an increase in their systems thinking skills when they learned in both a computational modeling condition and a paper and pencil‐based modeling, however the former had a larger effect size than the latter.…”

“…Research has suggested that fluency and flexibility of using different forms of model in a scientific explanation, called representational competence, show possession of deep conceptual understanding (Anderson et al, 2013; Gilbert & Treagust, 2009; Pande & Chandrasekharan, 2017). In addition, engaging students in any types of modeling activities have been found to support the development of advanced thinking skills, such as systems thinking (Nguyen & Santagata, 2021), which is among essential crosscutting concepts listed in the Next Generation Science Standards (NGSS Lead States, 2013).…”

“…Recently, modeling activities have been shifting from traditional modeling using physical models or drawing toward more computer‐based activities (Coll & Lajium, 2011). Emergent research has now suggested engaging K–12 students in computationally rich science activities (e.g., when students work to build scientific models through computer programming) to help enrich students' abilities to create representations of scientific concepts (Moore et al, 2020) as well as systems thinking (Mahaffy et al, 2019; Nguyen & Santagata, 2021; Scherr & Robertson, 2015). Such activities have given rise to new symbols, signs, codes, and even languages normally associated with the computer science (CS) discipline and used by students to understand scientific phenomena (Abrahamson & Sánchez‐García, 2016; Aksit & Wiebe, 2020; Chandrasekharan & Nersessian, 2015).…”

Transforming Everyday Information into Practical Analytics with Crowdsourced Assessment Tasks

“…A score of 1 would indicate some non-linear links, but incomplete evidence. The rubric was developed based on frameworks of science argumentation [34]. Prior to the crowdsourcing task, the second author and a research assistant coded 25% of the student responses to establish inter-rater reliability for the rubric: Cohen's κ = .73 for Elements and .88 for Coherence.…”