Thank you for your interest! We do offer some CS/CT training depending on the length of classroom implementation (from 45-min tasks to a longer, multi-task training unit); however, our curriculum itself is scaffolded into Inquiry, Instructional, Model Building and Challenge tasks. One of the aims of this scaffolding is to help with initial difficulties in CT/using the technology.

Inquiry tasks provide pre-built models and students are required to explore models by adjusting parameters. Here is an example for learning about 1D Acceleration. Instructional tasks are highly scaffolded and focus on the use of the DSML blocks to dive deeper into the target STEM domain (example built model). Model Building and then Challenge tasks require more CT applications (e.g., coding a medical delivery truck to speed up to a speed limit, maintain the speed, and then slow down to a complete stop at a stop sign). Finally, we do have embedded assessments that help us monitor gains in CT and the STEM domain over the course of the curriculum. We are constantly working to make that initial CT/CS introduction more effective for all and these formative results have supported this process.  

Great question, Abby! There are instructional experiences designed for getting students off the blocks with programming and becoming familiar with the programming environment before they delve into the Science modeling curricular activities. There is also a need for students to understand the step-by-step modeling paradigm (where the variables get updated with each time tick), and to understand the specific pre-built blocks (like 'simulation step') that they need to use. Across our various studies (and as part of our design-based research), we've been tweaking / refining the duration as well as the design of those activities focused on helping students learn programming. Of course, familiarity with the platform and technology is honed as they create their own programs/models.
The curriculum essentially aims to have students building their own models by the end of curricular units. We incorporate "guided inquiry" activities at the beginning to scaffold the initial learning process. We've tried a few different things in our different studies, and yes, this could involve experimenting with a pre-built model to understand the variables and relationships. However there are pre-built "domain specific modeling" blocks that essentially create a scientific "modeling language" specific to the topic they are learning.
Lots of thought and discussions in our project meetings around the exact nature of these inquiry activities to help students understand the platform and how to approach the task of building models :-)

[Didn't see Nicole's response before I started typing.. I guess we were typing around the same time :-)]

What a neat way to scaffold while still integrating! I enjoyed checking out the 1D Acceleration. 

Shuchi, do you have any plans to implement "ipsative assessments" in the curriculum? ,)

Here's a resource I recently read that elaborates on ipsative assessment (the section on "assessment as learning"):

Manitoba Education. (2006). Rethinking Classroom Assessment with Purpose in Mind. Manitoba Education, Citizenship and Youth.

Thanks for your query and for sharing the link, Jared! I look forward to reading it.
Although we have not formally planned on ipsative assessments, our embedded assessments do have that feel. Will certainly look to expanding/reorganizing their role in the learning process.

Great to hear!

We use the term synergistic learning to refer to the simultaneous learning of both STEM and CT. When we say collaborative we are referring to the students working together on a task. Thanks for the clarifying question!

C2STEM works like Google Docs: students can work on a joint project from their own computers.

Hi Michael! Optimally it is complimentary to other teacher-created activities, but it depends. We work closely with teachers to best adapt/design to their needs. For instance, in this video, our Physics teacher utilized our computational modeling tasks as a lab component, replacing traditional lab activities (completed by the control group) with C2STEM. Students received similar lectures - although slightly altered to incorporate information from the computational models/labs - and completed his classroom assessments as well as a few other handout activities. Occasionally, students completed the more scaffolded instructional tasks for homework to provide more in-class time for lectures/other activities (completed by both groups). The middle school science teacher utilized approximately 5 school days for marine biology tasks as part of a larger unit and incorporated video and photo discussions to connect the models to things the students learned in class. On the opposite end of the spectrum, we have implemented shorter studies in which the teacher's role is mainly a management role with C2STEM as the dominant mode of instruction. Thank you for the important question and I would love to hear how this approach may work best for your classroom! I can also be contacted at nicole.m.hutchins@vanderbilt.edu.

Hi Michael, To add to what Nicole described, we've also had high school Physics teachers who've chosen to use the C2STEM computational modeling units as a weeklong review unit after completing their traditional instruction. In such scenarios, many/most students are familiar with the Physics content, so the focus is on helping them model the Physics relations using a new representation - a computational one, and fostering a discrete step-by-step way of thinking and reasoning about the Physics relations using this computational representation.

Great question, Lisa! As a former high school CS teacher (with a non-CS undergraduate degree), this is a critical component for me! Previous CS knowledge/experience is not needed (a few did not have experience), but it does help - especially experience with block-based programming environments. In the lead up to implementation, we work closely with teachers to mold the implementation based on what the students have or will be learning in each class. During this time, teachers complete tasks as students and we communicate closely with them so they are comfortable giving instructions and leading discussions using the computational models and component tools. We are fortunate to have completed quite a number of studies thus far. As such, we are able to develop and distribute teacher PD packets that include learning objectives (in STEM and CT mapped to key standards), potential discussion topics, student difficulties experienced with similar tasks and more depending on the needs of the teacher.

C2STEM is a relatively new environment; however, as part of the development team I know we absolutely consider design capabilities and requirements for scaling up. For instance, while the domain-specific modeling language (DSML) blocks have proven beneficial for student learning in STEM+CT, the ability to customize blocks for specific purposes (for instance, DSML blocks to support ESL learners) has come up as we aim to broaden participation in computational modeling. Also, our development team is in the process of designing a teacher dashboard in which we aim to include more resources and curriculum development tools that we believe will support the scaling up process! Let me tag in other team members here as they may have more on this topic! 

Yes! I loved your video and am a fan of the visualizations in your modeling environment (and of course your methods and curriculum)! Did you start off with a BBPE such as Blockly / do you have recommendations regarding the visual design of those coral simulations?

We have also noticed the need to plan for differences in prior knowledge in programming. In current versions, we have created challenge problems that teachers do not often require of students (the model building tasks discussed in a previous comment also have more difficult CT applications) and these have served as a venue for those more advanced programming students. But we have also discussed the need to provide additional resources or support (potentially adaptive feedback during model building) at the earlier stages for those with little to no programming background. We have found this could be especially useful for issues such as better understanding of variables, the role of delta t, and the separation of the initialization process and the updating of the object's behavior. Dr. Grover has been very influential in better understanding and planning how we may target these differences in CT/CS prior knowledge and she'll be better able to add to this (busy meeting day, though!). 

Thanks for your comments, Ari. I too enjoyed watching your video and learning about your approach, especially given the overlap in science context (marine biology and coral reefs.)
Your question on differentiation is a great one. The short answer would be that we have not yet nailed this challenge completely - it's one for which we don't have all the answers but we have been thinking about it a fair bit and have tried a few different things in our studies.
To add to Nicole's comment-- we've run about 5-6 different studies in classrooms with students who had varied levels of programming experience (as a class, and even individually). A couple of observations related to this -- (a) given the increasing efforts to introduce coding to students, several of our students already had some exposure to programming especially in block-based programming environments. One of our motivations to use the NetBlox extension of Snap! was the familiarity many students now have with the Scratch/Snap!-kind of coding environment. (b) Even when students had knowledge of programming, their prior experiences did not necessarily translate to an intuitive understanding of how to create computational models. We found that all students need to understand specific aspects of the model-building process - identifying variables, their initialization, the relationships between variables and how to express those in code, how variables get updated in each time tick (the iterative calculus), stopping conditions, and debugging their models when they do not work as desired. And each of these interestingly combines learning of both Physics and coding (or what we call "synergistic learning"). Our ongoing plans for upcoming studies includes thinking more closely about this breakdown into subgoals, and inquiry strategies around them.
Our curriculum takes a scaffolding/inquiry approach that gives students an introduction to programming (the needs there vary by prior experience, and we've been refining our coding training to be more relevant/effective for the model-building tasks ahead) and then guides them through model-building tasks (set in real-world contexts), and finally leads to challenge tasks where students must code a model without scaffolds. Although it's still something we're still iterating on, we've found this approach to work for most students. Additionally, some of our studies have involved students working in pairs or groups of three -- collaboration in such settings helps students in their thinking and problem-solving.

Our research studies have thus far also involved planning for additional challenge tasks for students who are farther along.

Hope this is helps somewhat.

Yes, thanks to both of you--super helpful comments! We've had very similar observations in the middle school environment--the integration of CT with science affords different entry points, and also means that different kinds of scaffolding strategies are needed (e.g. for science concepts & practices, for CT/CS concepts & practices). 

Love this thread--and your project! As Ari indicates, this sort of CS/S domain bootstrapping idea is something we're really trying to figure out. Toward that end, we're also wrestling with the measurement side of it: the "synergistic learning" concept really resonates. Have you been approaching it as a combination of independent CS and S metrics, and/or do you have an integrated measure of it? I got the sense from your video that you share our wonderings about whether something special is happening with the learning at this intersection (e.g., do kids understand the underlying science more deeply if they need to engage computationally with a scientific model?)

Can't wait to learn more about your work!

Hi Eric, Thank-you for your comments. For this project, we've approached the measurement of synergistic learning using a multi-pronged approach. We've been using independent CS/CT and science measures for summative purposes, and integrated measures for formative purposes. The integrated measures involve formative assessment tasks or what we call "check-ins" where we have students using a given computational model to answer a question, or debugging, developing or modifying given computational models of kinematics phenomena. In addition, we also have a post-intervention Preparation for Future Learning assessment to measure how well students can transfer this discrete step-by-step way of thinking that we encourage in C2STEM to solve other problems in other domains and contexts.

How have you been dealing with the measurement issue in your project? Btw, I really enjoyed your video and would love to know more about your project.

Thanks, that's super helpful! 

For our project, we have been piloting measures this year to capture science/CS learning as well as dispositional constructs pre/post (see below). This is in preparation for a broader study of the coding science curriculum next year. We've been focusing on:

• Valuing computer science (adapted from the Science Learning Activation scales)
• Competency beliefs (also adapted from the Activation Lab suite of measures)
• Identify (e.g. I am a coder)
• Disposition to pursue CS/STEM

Our computer programming/computational thinking measure builds on measurement development work by Chris Schunn's team at Pitt to create middle school friendly CT scales that draw from item sets in SRI's PACT work. 

As part of another study (DRL#1838992), we are developing CT-science integrated tasks that we expect to be able to use as a pre/post measure in next year's study. Items from the CT-STEM project at Northwestern have been helpful in that effort.

Thanks, Eric! I enjoyed watching the video of your project - and can see how your team may be wrestling with some of the same questions.

One of the best sources for insights into the synergistic learning (that special something, as you put it) that we have so far is the collaborative discourse as groups (of 2 or 3) students work together to build their models. The collaboration and conversations help externalize their thinking. We found some aspects of the model-building process to be more conducive to foster the learning at that intersection of CS & S.

You can see some of our preliminary findings here (https://www.researchgate.net/publication/332427... & https://www.researchgate.net/publication/329553...
We also have a paper on this (Synder et al., 2019) at the upcoming CSCL conference.

Looks like our messages crossed, Eric.. Was responding to your earlier message!

awesome--thanks!

and yes, we've noticed similar things in student collaborative discourse, as students work together to wrestle with the code and the science of a model--it's been really exciting to see how the two domains come together in those moments. Looking forward to checking out your papers!

ha, and they crossed again, it seems...

Thanks for sharing the measures you've been using! Super helpful to know how others are measuring CT & Science competencies.

Some of our CT measures also come from the PACT work, and others from a subsequent project (VELA) in middle school classrooms that focus especially on variables and expressions (a key aspect of computational/science modeling).

Thank you very much! 

Thank you, Michael! And that is a great idea. As we push forward in expanding our reach (grade levels, domains, broadening participation of underrepresented minorities and women), we have definitely witnessed the need for multiple approaches to the curriculum and techniques for orchestrating the classroom environment. This idea could also be great in creating a community that could be a resource for inclusive activities and a forum to discuss challenges. We are currently working on a teacher dashboard that could include a library of these techniques (in addition to other PD material). We have also recently set up expert.C2STEM.org as a preliminary forum, but I will definitely bring this idea to the team!