Great question about the transitions between science and programming. In our most recent classroom study, we actually had kids moving the opposite direction— they began by tinkering with the sensors, actuators and programming environment, then transitioning into scientific investigations. The idea there was to familiarize and engage them with the tools and software, then have them use the (now more familiar) tools to do scientific investigations.

In this “tinkering” stage in particular is where we saw lots of computational thinking (which I’m using here to mean reasoning about the behavior of computational systems). Even prior to science content, students made sense of sensor noise, delay times, or the effects of threshold values set in their program on the behavior of the sensor-controller-actuator systems they built. When these same tools are reused to do scientific investigations, we see opportunities for scientific *and* computational thinking, for example when a student reasons about their data, teasing apart sensor signals into biological and computational components (ex: signal and noise).

As to your question, “In what ways are students experiencing computational thinking?”— this is one of our primary research questions! We’re designing for CT both in “tinkering” activities, but also are designing into Dataflow a way of thinking about how computational systems act on data: by using the software to collect data at different rates, transform data, store data, and act on data (with the actuators), we hope to make these computational concepts and processes visible, inspectable and manipulable by students. However, how students *experience* this is an open question for research!

Are they formulating problems and solutions that leverage computational resources? Yes! For example, we have designed activities where students use a relay, lamp and control program to stabilize the level of CO2 in a container with spinach leaves (photosynthesis drives the CO2 down when the lamp is on, and cellular respiration drives it back up when the lamp is off). This sort of design/control activity gives students opportunities to reason about the behavior of this system using both biological and computational concepts.

Thank you for your question Brian. We are deeply focused on the computational part of our STEM+C project at the moment. We're very excited to get to see students working with sensors in more complex systems over time. Our first thoughts were along the lines of "bottle biology" and/or plant chambers depending on teacher comfort level and available classroom time. We'd love to hear any experience you or others might have at the 9-12 level with longer term ecosystem projects, plants, bacteria or otherwise.

Great question, Brian. I want to follow up on Tom's answer. Right now we are working to design the tools to support investigation at many different scales - from one or two days activities that educators can fit into existing curricula, to multi-week units that move from "messing around" with sensors and relays, to independent investigations with learner-designed microcosms. In pilot studies, we have been working through both of those scenarios. In the med/long-term implementations, students are indeed asking their own questions - in some cases about specific processes of photosynthesis and respiration, in other cases about more complex interactions of an ecocolumn.  Glad you asked about even longer term too - one thing we are excited about is being able to use the open source actuators to create closed (or almost closed) systems into which students can build their predictions. These could run for months, and a couple already have. As we extend the software, we are also hoping to be able to allow learners to collect data in schools, homes, and communities, and then put these datasets in conversation with what they are seeing and doing in the microcosms. We're hoping that the ability to move across scales in this way will be another to create multi-year studies and datasets. 

Thanks Brian for checking out our videos and the idea of the Winogradsky column. They are so beautiful once grown even if a bit smelly. One design issue that we have been struggling with in this project is the desire by teachers to have units fit into the schedule and structure of the school day while real biological phenomena (with exception to yeast and some molds) are slow slow and require longer term observations. As Tom mentioned, this is a topic we are fascinating with.

Hi Katie! Right now, we are analyzing primarily interview and observation data, with a focus on students' understanding of "where data comes from," when and how they tinker with and interrogate their computational systems, and what this may mean about their relationships to data and data production.
As we refine our understanding of where students can (and want to) take these tools, we’ll be developing instruments to better describe and measure this.

We have seen students coming to attend to confounding effects on their data by the end of their investigations, or adjusting their thinking about the relation between their data, instruments and procedures, and conceptual models. We would argue these are evidence of a shift in their understanding of the nature of that data (and, through many such experiences, could contribute to a shift in their relationship to data more broadly).

This is all very interesting to me, thank you for your presentation. To follow up on what Katie asked, could you say more about what constructs you think you will measure eventually? Do you currently have some sort of model or theory of change that you're expecting to come into play, and is there a set of specific outcomes you're hoping for for these students in terms of their understanding of the data?

We are taking a design-based research approach where we are concurrently designing materials, activities, and a pedagogical approach with design conjectures about learning, while refining our theories about how students learn, so it is a bit early to say what constructs or specific measures (in the educational measurement sense) that we will have here. We are interested in understanding to what extent and under what conditions are students able to use the project’s computational resources to undertake authentic scientific investigations. How do students develop agency over scientific data? What changes do we see in students when they engage in this curricular approach that integrates science practices with computational thinking practices? What kinds of background materials and assistance do teachers require to effectively support the intended activities? We of course would like to see improvements on all fronts for what students learn in biology class, but value practices that can serve them in the long run such as helping students engage in sense-making with data, taking up practices like being intentional in their choices and being more systematic in designing and revising experiments, tinkering with lab tools to harness data from phenomena, interrogating and representing the data (from their ecochambers, Ziplock bags of kale, or other gas-producing/consuming plant cells.)  Working with digital data (collected from their IoT sensors) is as important as developing “thing knowledge” such as knowing how and working with data measurement tools and physical instruments. As an aside, we are leveraging Concord Consortium’s open data analysis platform called CODAP for students to use for further data analyses. We are just under 20 months into our project, so it’s a busy, but very exciting time in our research. Feel free to ask more questions.

Thanks for your question Midge. In our work, we haven't positioned our work to examine cognitive transfer per se, but perhaps could be considered the “methods” that you refer to? We hope that students will develop a set of practices that will be useful and empower them to answer their own questions. We are interested in looking at the kinds of practices that students engage such as those found in three-dimensional learning from NGSS. These include asking questions, making sense of data that they produce themselves from exploring a biological phenomenon, planning and carrying out a scientific investigation with authentic measurement tools, and others. In the design of InSPECT activities, we have been thinking hard about how to support student agency in scientific investigation where computational tools are involved. (I will check out your Habitat Fragmentation module. Sounds interesting.)

Hi Midge! Building on Sherry’s response— the transfer question is a great one. We have discussed this internally, though transfer per se has not been a primary focus of our research. However, we have been focusing recently on a closely related issue-- of when students bring prior knowledge and experience to bear on scientific investigations with these computational tools. In line with constructivist learning approaches, and with the goal of supporting students' identity and developing agency within science, we aim to design for this.

As a related anecdote, a student told me today that after doing our "Spinach Lab" (in which they observe CO2 levels in closed containers of spinach, where the CO2 levels rise in the dark due to cellular respiration), that she now wonders whether her refrigerator lets out large amounts of CO2 when she opens the door! This isn't a classic example of transfer, but we do see instances like this as evidence of students connecting their in-class activity to knowledge and experience outside the classroom.

Sherry and Lisa-Thanks for your video AND your continued discussion within the comments.   I might have missed the very encouraging comments and the fact that the focused activities that your 9^th grade bio video shares seem like such a natural flow from where are project’s physical /EarthSpace Science activities carry  6-8^th graders.   Fortunate to have the inconnectivity of systems thinking of the STEM-for-All Showcase to encourage us as researchers to view other videos and truly think STEM as a whole.    Also reading the struggle to show what you are finding from one grade and application in an ongoing form as the students advance seems just a manner of time.  I checked in here to learn more about the sensor progression and learning more than I thought.  Must be that we are truly adjusting to all the changes in this era and able to watch the systems form.

 We found the Energy3D developed through Concord Consortium and just beginning to apply it to our connections with the new content with USGreen Building Council.  Fits in an “inside out” form from your labs and starting the exploration with the CT thinking of placing the sensors. The challenge of encouraging students to use their whole classroom as their lab prepares them to enter your 9^th grade lab experience with a shared background. (see the end of   http://stemforall2018.videohall.com/p/1321)   Just a matter of time until these pieces form a flow of students.  Congrats on your exciting progress and parts that help the rest of us tag along. 

Thanks Betsy for sharing. As we hear from different teachers in our own work, students do get a wide range of science preparation in the middle grades before showing up to high school, so I'm grateful to learn more about efforts like STEM Literacy Community of Practices from your video. I was also fascinated to learn about the online personalized labs and Inq-ITS that score themselves and provide individual formative feedback. It could be a nice progression and complement to the hands-on team labs encountered in InSPECT labs in 9th and 10th grade biology.

Thanks for noticing! The video was produced locally by Tiny Oak Media in Oakland. We did not tell the kids what they could or couldn't wear to school.  (The school's logo happens to be printed on polo shirts, black sweatshirts, and hoodies.)

Thank you, Emerlyn! We are glad (and so humbled) by your comment. It is our hope that we will be able to broaden participation for students working with data.