Never far from shore: productive patterns in physics students’ use of the digital learning environment Algodoo

In this paper, we present three types of activity that we have observed during students’ free exploration of a software called Algodoo , which allows students to explore a range of physics phenomena within the same digital learning environment. We discuss how, by responding to any of the three activity types we identify in the students’ use of Algodoo, physics teachers can springboard into a range of relevant physics discussions while supporting and valuing student agency and divergent thinking. Thus, while one might not expect students’ undirected use of a digital tool such as Algodoo to be particularly worthwhile for the physics classroom, we highlight how students are never ‘far from the shore’ of a productive physics discussion.


Introduction
With regards to digital learning environments in physics education, on one hand there are software that focus on specific phenomena, e.g.
Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
PhET simulations [1], Physlets [2], QuVis animations [3], and, on the other hand, software that function more as creative arenas within which many phenomena can be explored, e.g. Algodoo [4], Interactive Physics [5], and Fizika [6]. We refer to the former category of phenomenaspecific digital learning environments as 'constrained' and the latter as 'less-constrained' [7], owing to the degree to which those environments are designed to 'productively constrain' [8] students' behavior (see also [9][10][11]). While both constrained and less-constrained digital learning environments have unique affordances for supporting physics students' learning, constrained software has received significantly more attention from developers and physics education researchers in recent decades. In this paper, we present our findings for some of the ways in which one less-constrained digital learning environmentspecifically, Algodoo-can be productive for the physics classroom, specifically when students are allowed to freely explore within the software for themselves.
Given the relative open-endedness of lessconstrained software such as Algodoo, a physics teacher who includes this type of digital learning environment in their repertoire of curricular materials can choose to implement the software in a range of ways. At one extreme, a teacher can use Algodoo as a means for students to engage with specific physics content in directed tasks. In this approach, students can use Algodoo to explore a range of physics topics, from projectile motion [12] to Kepler's laws [13], kinetic gas theory [14], and the refraction of light [15], all within the same digital learning environment. This topic-specific use of a less-constrained software more closely resembles the productively controlled approach behind many constrained physics learning software mentioned above 3 .
At the opposite extreme, a physics teacher can choose to take a more student-directed approach: that is, a teacher can refrain from selecting specific topics, allowing students to explore the software for themselves and responding to the students' exploration at opportune points. Such a student-directed approach may intuitively seem too unfocused to be worthwhile for the teaching and learning of physics. However, we have found that students' self-directed exploration within the less-constrained software, Algodoo, has several unique, if unanticipated, affordances for physics education [7].
In this paper, we describe three types of activities identified while observing students' selfdirected use of Algodoo and explain how each activity type has the potential to be productively leveraged by a physics teacher. Among other things, we show that, by allowing students to creatively explore within tool-rich physics environments such as Algodoo, physics teachers can springboard into a range of relevant discussions while supporting and valuing the agency and divergent thinking of students.

Three types of student activity in Algodoo
The three activity types we present in this paper were identified based on the analysis of video recordings of seven pairs of university physics students as they used Algodoo for the first time 4 . Through the use of a grounded theory method [16], the video recordings were iteratively viewed, transcripts were generated, student behavior was coded in successively larger chunks, and, ultimately, three categories of activity were identified. Below, we present each of these activity types, along with their potential relevance for the teaching and learning of physics.

Activity type 1: exploration of the software fundamentals
During the first activity type, which we call exploration of the software fundamentals, students investigate the tools and functions of Algodoo. This activity type is characterized by students familiarizing themselves with Algodoo's buttons, toolbars, and drop-down menus (see figure 1) in order to develop a sense for the basics of how the software is operated. Students' exploration of the software fundamentals tends to be (outwardly) chaotic, exemplified by the students shifting their focus between the many features housed in Algodoo. Especially when students are new users of Algodoo, this activity type is the behavior which physics teachers 4 The activity types presented in this paper were originally identified in Prytz's master's thesis [28]. For those interested, a full discussion of the data and methodology can be found there. should expect to see chronologically first. Additionally, due to the high number of features made available in Algodoo, students frequently return to this activity type throughout their exploration as they discover new functionalities of the software.
Students' exploration of the software fundamentals can be productive for the physics teacher insofar as students naturally uncover new physics parameters. While students 'poke around' in the Algodoo software, they interact with the buttons and sliders corresponding to various parameters that are relevant to the discipline of physics (e.g. restitution, damping, speed/velocity, gravity, kinetic energy, and so on). Depending on the students' familiarity with the formalisms of physics, the parameters encountered in Algodoo will be more or less recognizable to the students. A physics teacher can notice which parameters seem to be less-recognized by students and encourage those students to further investigate those parameters unfamiliar to them. For example, elsewhere we have analyzed how two students came across the slider labelled 'Damping' within the editing menu for springs (figure 1, right) and were guided by researchers to make sense of the behavior of damped/undamped springs [7].

Activity type 2: testing and contrasting
In the second activity type, which we call testing and contrasting, students explore how well the physics engine of Algodoo matches the real world. Students tend to construct classic physics scenarios within the software to ensure that the software behaves as it should (e.g. dropping two objects from the same height to 'demonstrate' the acceleration due to gravity) or they create simple tests to determine if the software allows for certain complexities of physics interactions (e.g. dropping a square onto a thin rectangle given the material preset of 'glass' to see if glass objects shatter in Algodoo 5 , figure 2). In general, we have found that students engage in testing and 5 Algodoo allows the user to assign material presets such as glass, wood, steel, etc, to objects, which correspond to certain configurations of the objects' density, friction properties, restitution, and refractive index.

Figure 2.
A screenshot of an observed testing and contrasting activity in Algodoo-recreated by the authors for clarity in this publication-where two students tried to shatter a thin rectangle made with the 'glass' material preset (shown in light blue laying horizontally on top of two support rectangles) with another, massive object (the gold square at the top of the screen). After running the simulation, the students concluded that Algodoo does not allow the breaking of objects.
contrasting to check 'how far' they can go within the software.
For the physics teacher, testing and contrasting activities can be productive in that they can be leveraged to highlight the role of modeling in physics [17,18]. For example, in the case shown in figure 2 involving the 'glass' rectangle, a physics teacher could interject to ask students about whether the Algodoo software is still a valid environment for exploring physics phenomena. Previous research findings have highlighted that software such as Algodoo have the potential to be a resource for entry-level physics modeling [4,19], where the intuitive interface allows students to examine dynamic models of physical phenomena without a prerequisite proficiency in programming. Beyond supporting discussions of physics modeling, students' testing and contrasting activities can serve as the backdrop for introductory discussions around the use of computer simulations in modern science [20].

Activity type 3: engineering
In the third activity type, which we refer to as engineering, students tend to prototype machines/setups in the pursuit of self-determined goals within Algodoo. For example, some students constructed a simple car and, after getting their car rolling, created obstacles such as a small hill for the car to traverse (figure 3). In doing so, these students were motivated to explore the role of friction, torque, etc, in the context of a car climbing a hill. In many cases, the students' impetus to construct machines comes during their noticing a specific feature of Algodoo in their exploration of the software fundamentals (activity type 1). After finding Algodoo's 'thruster' tool, for instance, students might quickly transition into building a rocket. We have found a salient feature of engineering activities is students' modification of their machines in pursuit of their self-determined goal. In this way, students' engineering activities seem July 2020 Figure 3. A screenshot of the kind of machine typically created by students in Algodoo-again, recreated by the authors for this publication-as they engage in the third activity type, engineering. A simple car consisting of a rectangular 'body' and two circular 'wheels' (fitted with motor-driven axels) drives up an angular obstacle also created by the students. In achieving the goal of driving their simple car up and over this obstacle, students engage with the coefficient of friction of the wheels/ramp and the torques applied to the motor-driven axels, among other things.
to involve iterative loops wherein they design, test, and challenge various prototypes.
Students' engineering activities are potentially useful for the physics classroom in that they often entail students working in ways that resemble pedagogically-sought-after scientific practices [21]. More precisely, during the prototyping typical in engineering activities, students often design tests for achieving self-determined goals, evaluate the outcome of their tests, and iteratively revise their constructions to accommodate those outcomes. This sequence of reasoning may constitute what Gregorcic et al [22] describe as the 'seed[s] of scientific practices'-or perhaps in our case, the seeds of engineering practices. A teacher can take advantage of students' engineering tasks by prompting students to reflect on their prototyping process, which may subsequently be turned into a discussion of the characteristics of scientific and engineering practices more broadly [23,24]. Alternatively, since Algodoo allows for the 'zooming in on' and the 'unpacking of' the various parts within a construction, teachers can utilize students' machines/setups as the focus for further inquiry. For example, when a pair of students engineered a way to lodge an arrow into a 'sponge' in Algodoo, researchers were able to utilize the students' success in meeting their goal as the basis for a discussion on how non-rigid bodies are modeled in the software [7].

Summary of implications for teaching
When allowed to explore software such as Algodoo in a self-directed manner, students are likely to engage in some combination of the three activity types detailed above, each of which can be potentially productive for the teaching and learning of physics in their own way (table 1). A teacher who allows students to engage in this kind of activity can choose to build upon students' exploration [25] during any of the three activity types Table 1. A summary of the three activity types we identified in students' use of Algodoo without a specific prompt.

Activity type Explanation
Productiveness for the physics classroom

Exploration of the software fundamentals
Students develop a sense for the tools and functions of the software environment Students can naturally uncover relevant physics parameters Testing and contrasting Students explore how the software's physics engine compares with the real world Can serve as a backdrop for discussions about physics modeling and the use of computers in science Engineering Students create machines/setups and pursue self-determined goals Students' engagement in science-and engineering-like practices may be unpacked for further discussion; students' creations can be taken up as the focus for further inquiry in order to guide the students' attention to the type of discussion that the teacher wants. Building from the first activity type, students' exploration of the software fundamentals, a teacher can encourage students to explore the meaning of various parameters as they are uncovered in the software interface. With students' testing and contrasting, the second activity type, a teacher can springboard into entry-level discussions on the role of modeling in physics and the function of computer modeling in scientific inquiry. Finally, from the third activity type, students' engineering, teachers can take on students' creations for discussions around scientific practices and/or as inspiration for more topic-specific discussions of particular phenomena. Among other things, allowing students to explore less-constrained physics software such as Algodoo in a self-directed manner can signal to those students that their creativity, divergent thinking, and, ultimately, agency [26] is valuable to the process of learning-and doingphysics.

Additional considerations for teachers encouraging students to explore Algodoo
While we have so far used this paper to highlight ways in which students' use Algodoo can be productive for the physics classroom, it is worth mentioning some potential challenges physics teachers might face when encouraging an entire class of students to 'freely explore' less-constrained digital learning environments. First, in larger classes it may be difficult to manage multiple groups of students that may be headed in divergent, often unrelated, directions. In our experience, we have found that a single teacher, who is familiar with the software, can manage classes on the scale of 20-30 students (in groups of 2-3). Alternatively, teachers can encourage students to explore Algodoo as homework or within a labstyle setting where student-teacher ratios tend to be smaller. A second challenge that physics teachers might face when implementing Algodoo in their classroom is that allowing students to 'mess about' in such digital learning environments simply takes more time to reach certain learning goals when compared to more pointed discussions/lectures of desired topics. Our recommendation regarding this concern is that teachers consider incorporating student-directed use of Algodoo into their toolbox of active learning approaches and, ultimately, decide for themselves what balance to strike in the pacing of content goals.
Finally, since every student group will, by design of such a teaching approach, end up engaging in different activities and pursing their own goals, a third challenge teachers might face when incorporating Algodoo in their classrooms is that there is a loss of a single shared experience for students. In response to this last challenge, we recommend that teachers intermittently pull the full class together for a discussion around a particular group's work. This may help all of the students reflect on the nature of their own work and may encourage them to generalize some of their productive ideas across the various contexts that appear in other students' work [27].

Conclusion
Regardless of whether students are 'poking around' in Algodoo's menus, checking the boundaries of the software's physics engine, or creating machines to meet their own goals, physics teachers can be assured that students' self-directed exploration of less-constrained digital learning environments is a near neighbor of a worthwhile physics discussion. Thus, as long as the interested physics teacher is willing to build upon the divergent activity of students engaging with lessconstrained digital learning environments such as Algodoo, their students will likely never be 'far from the shore' of productive physics discussions.