Mechanistic reasoning in science education: A literature review

Abstract

There is a growing research interest in mechanistic reasoning (MR) in the field of science education, as this type of reasoning is perceived as an essential thinking skill for science education. This literature review synthesized 60 science education studies on MR published from 2006 to 2021. The findings showed three common aspects of conceptualizations of MR in science education: (1) causality in relation to MR, (2) use of entities and their associated activities, and (3) use of entities at (at least) one scale level below the scale level of a target phenomenon. While most of the reviewed studies related the importance of MR to cognitive aspects, a smaller number associated its value with scientific modelling. Three main difficulties in generating MR were categorized: (1) identifying and using unobservable entities, (2) assigning activities to entities, and (3) identifying and using an appropriate number of entities. Various types of support for fostering MR were identified. Implications and future studies are discussed.

When a machine detects student reasoning: a review of machine learning-based formative assessment of mechanistic reasoning

“…In their dimension construct, Zhai et al (2020a) found that 81% of the reviewed studies tapped complex constructs indicating the great potential of ML in assessing such constructs. Since mechanistic reasoning is a complex construct per se (cf., Bachtiar et al, 2022), our review confirms Zhai et al's (2020a) finding. However, in many studies of our sample, there is no evidence that the investigated construct could solely be analyzed by ML.…”

“…Across all sciences, mechanistic reasoning is a critical thinking skill (Machamer et al, 2000;Glennan, 2002;Russ et al, 2008;Bolger et al, 2012;Illari and Williamson, 2012;Southard et al, 2016;van Mil et al, 2016). However, students experience significant challenges in generating mechanistic explanations in science subjects (cf., Bachtiar et al, 2022). Research in chemistry education has as well documented that these challenges include reasoning about chemical mechanisms in organic (cf., Graulich, 2015;Watts, 2022a, 2022b) or general chemistry (cf., Talanquer, 2009;Sevian and Talanquer, 2014) since many cognitive operations must co-occur.…”

Understanding how student‐constructed stop‐motion animations promote mechanistic reasoning: A theoretical framework and empirical evidence

“…Highlighted by our literature study (Bachtiar et al, 2022), also reported by other studies (e.g., Andrade et al, 2021; Bachtiar et al, 2021; Wilkerson et al, 2018; Wilkerson‐Jerde et al, 2015), students used MR when reasoning with their own creation of a model of a phenomenon. Given a growing research interest in student‐constructed stop‐motion animations (SMA) as digital explanations that have been shown to support science learning (see Farrokhnia et al, 2020; Hoban, 2020; Nielsen et al, 2022; Paige et al, 2016; Yaseen & Aubusson, 2018), our present study focuses on exploring the contribution of student‐constructed SMA for specifically supporting MR.…”

“…The findings in our current study add the knowledge to the field by emphasizing that students' engaging in constructing chunks and sequencing these chunks (by thinking about the transition between the sequential chunks), led students to employ the basic elements of MR, that is, identifying entities and activities of entities. It is precisely these elements that were found to be the most challenging elements that students need to involve in when developing MR (Bachtiar et al, 2022).…”

“…We posit that constructing this kind of explanation needs to use a type of thinking called mechanistic reasoning (MR), which specifies both entities and the actions these entities take. In addition to the benefits of MR to generate scientific explanations, a recent literature study shows the importance of MR in science education (Bachtiar et al, 2022), also highlighted by science education research (e.g., Balabanoff et al, 2020; Becker et al, 2016; Macrie‐Shuck & Talanquer, 2020; Russ et al, 2008; Southard et al, 2016; Weinrich & Talanquer, 2016; Wilkerson et al, 2018). For example, students who used MR to explain a physical phenomenon demonstrated a higher level of conceptual understanding than students not using MR (Balabanoff et al, 2020; Cooper et al, 2016; Crandell et al, 2019).…”