TrafficJams: Social Endeavors Within a Constrained Responsive Environment to Investigate Driving and Traffic

• In Event: Discovery-Based STEM Learning 2.0: Are We There Yet?

Abstract

Theoretical Approach
Discovery learning proponents designs that support active knowledge construction (Bruner, 1961) including open-ended problem-based, inquiry-based and design-based learning, showing advantages to learning and especially its transfer (Author, 2008). Opponents demonstrate how unguided discovery is ineffective, based on cognitive load and expertise (Mayer, 2004; Kirchner et al., 2006).

Objective
This preliminary design-experiment approaches this quandary by investigating learning with a participatory-simulation (Resnick & Wilensky, 1998; Colella, 2000) that supports open-ended collaborative endeavors within responsive constrained environments, by planning, enacting and interpreting social experiments. Collaborative endeavors, distinct from design or inquiry, widen the scope of activities that can take place. Constraints limit the range of possible actions. Responsiveness relates to feedback to effected actions. Pairing of environmental responses to constrained action provides for discovery of underlying rules, amplified through social interactions.

Methods
Design involved analysis of traffic, its paradoxes and misunderstandings by drivers (Kerner, 2004), resulting in goals: making such paradoxes apparent in the system’s responses and supporting enactment of a wide range of endeavors. Participants use the TrafficJams simulation (Authors, 2014c, see Figure 10 & 11) driving collaboratively on a two-lane road (constraints). In the process they discover the relationships between driving and traffic. TrafficJams was created with NetLogo HubNet (Wilensky & Stroup, 1999), based on Traffic (Wilensky, 1997).
Learning of one high-school class was studied with a pretest-intervention-posttest design during 2.5 hours. 29 students participated, 16 completed both tests.

Identical pre- and post-assessments (“self-aggressive driving”, Sikron, Baron-Epel & Linn, 2007; locally-created “understanding traffic”), observations and data-logging.

Results
After free driving to gain experience, the first challenge was making traffic jam. Students proposed three strategies, voted and two were enacted: driving closely and frequent lane changes. They chose to include accidents, which resulted in much gore and joy. Smoothing traffic was the next challenge and students designed two strategies: All cars in one lane, and having everybody drive at the same speed. Discussions included how an individual could impact the collective and vice versa, frustration at not being first overriding willingness to cooperate, concluding: “The point is that the same behavior of keeping your distance also quickens traffic and also prevents accidents.”
The questionnaires’ analysis show that most but not all students advanced their understanding of traffic for at least one dimension, mainly regarding complex-systems related issues of decentralization, distributions of causes and speed (Table 1). We see a shift from global “I think the jam is caused from merging a number of lanes …” to emergent-decentralized “Cars slow down and speed up during a traffic jam. I think this happens because not all the drivers drive at the same speed and that causes everybody to slow down, which causes the jam.” In the conference, data-logging and “self-aggressive driving” results will be presented.

Significance
Results support the claim for open-ended learning that is supported by a constraining environment’s responsiveness and collaboration, and suggest ways to enhance such learning in future experiments. This work aims to heighten safety in driving and create a culture where congestion works optimally.

Authors

• Sharona T. Levy, University of Haifa

• Ran Peleg, University of Haifa

• Eyal Ofeck, University of Haifa

• Ilana Dubovi, University of Haifa

• Naamit Tabor, University of Haifa

• Shiri Bluestein, University of Haifa

• Hadar Ben-Zur, University of Haifa