Hi Gerad, thank you for your question. We employ a mixed methods approach to collecting and analyzing student discourse and activity. These include pre and post surveys which measure growth in student's science content knowledge and ideas about modeling, as well as affective dimensions such as identity and feelings of competency related to science and programming. We also gather information on students' prior programming experience, as well as collect back end data during tool use to help us put together individual and aggregate-level trajectories for students and classrooms in terms of their tool explorations and programming decisions. We also conduct exit interviews with all teachers and collect daily feedback on each lesson. Each of these pieces of data help us assess the efficacy of our curriculum. 

Regarding your second question, in terms of design one of the biggest challenges are the curricular and institutional constraints placed on teachers - which often results in limited time each day and each week to teach science and tight restrictions on what can be taught - and how those constraints interact with our desire to support inquiry-based approaches to science learning. For this reason, we've made strong efforts to anchor the design of our curriculum in practice-based approaches to science learning, while at the same time designing for real classrooms. For example, all of our lessons are between 45 and 50 minutes and are closely aligned to both state and national science standards (e.g NGSS) as well as several standards in mathematics and digital literacy. There are, of course, other difficulties associated with implementing this type of curriculum, including supporting teachers in teaching programming. We have designed professional development tools and resources for this purpose, and have received good ideas for how to continue to support teachers in practice from our veteran EcoMOD teachers. 

Thanks for your response, Amanda. This is a really exciting project!

Your point about being adaptable to meet the realities of classrooms really resonated with me. Can you share a little bit about the general structure of a lesson? For example, how much of each lesson is spent using the simulation vs. group work or discussion? Can that 45-50 minutes be broken up across multiple class periods?

Thanks!

We think the project is pretty exciting, as well! Thank you for the kind words. 

In general, roughly half of the class is spent exploring the immersive tools or the programming environment, and the other half of the class is a combination of physical activities (for example, embodied modeling activities), sharing ideas or presenting and sharing programs. Each of our lessons also begin with a "Do Now" activity relevant to the day's goals, which has the added benefit of occupying students while the program loads. Depending on the strength of the school's wifi, our program can load in as little as a few seconds or as long as a couple of minutes. We didn't want to waste this time so designed brief orienting activities to start each lesson. 

This, of course, is a rough estimate. The actual time on each task varies as teachers adapt in the moment to student ideas. And yes, I believe the 45-50 minutes can be broken up into smaller chunks and we have observed some teachers do just that if circumstances prevent completing the lesson during the allotted timeframe.

Hi Bess, thank you for your question. EcoMOD is a stand-alone curriculum but has been designed to align with both state and national science standards, as well as several standards in mathematics and digital literacy. We designed EcoMOD in this way so that we can support teachers across diverse school districts, which often have different requirements for science teaching and learning. This also shifts agency to the school and teachers, who can decide how best to integrate the curriculum.

We do not have complete data on whether or not the unit was taught in addition to or in place of other curriculum, but it appears to be a mixture of both. In one of our schools, the district only requires two units on science (magnets and fossils), so EcoMOD is incorporated as a means to enrich the relatively thin science requirements. In other schools, EcoMOD is taught as an extension curriculum to prior lessons on ecology or as a final end-of-year unit after all other required curricula has been taught. 

Hi Feng, thank you for your question. The research questions investigated on the current grant include two questions on student learning (essentially, 1. what knowledge do student's gain 2. through what scientific or computational practices?) and two questions on teacher use and attitudes towards the curriculum. Given our questions, the natural variation between schools was not something we accounted for, however, as we move into next year we are considering conducting comparison studies where we will would account for some of the variables you mentioned.

Re: state standardized assessments. No, we have not looked at student achievement on those assessments.

Thank you for the supportive words, Nancy! You've hit on one of our big discussion points as we designed our curriculum - how many animals or variables will students be able to program or manipulate? Given the high variability in student prior programming experience (some classrooms might have all novice programmers, while others might have greater than 50% that have programmed before), we intentionally limited the animals that students could program to one or two (beaver and woodpecker). As soon as you increase the complexity of the system, in terms of number of animals students can manipulate, the difficulty of the modeling task increases. Based on our classroom studies, the level of programming difficulty is about what it should be given the length of the curriculum. 

Although students are only programming, for example, a beaver, non-programmable animals like fish and water fowl are also present in the model and students can observe and interpret data on those animals to see how they are impacted by the beaver. Additionally, students can add a predator to the system (and code predator avoidance behavior) as well as manipulate set up variables such as water flow, tree density, tree location and width of the stream to observe how those variables affect the beaver, the environment, and the other animals in the system.