FLORENCE SULLIVAN
Microgenetic Learning Analytics
NSF Awards: 1252350
2018 (see original presentation & discussion)
Grades 6-8
This exploratory case study is devoted to examining the development of computational thinking and doing in a one-day, all-girl robotics workshop. The aim of the study is to create knowledge related to how educational robotics can support the development of computational thinking in middle school aged girls. We collected audio and video tape data from six groups of girls (N=17) over the course of the one-day workshop. Here we present one aspect of our case study findings related to one group's development of progressively sophisticated strategies for solving a robotics problem. We demonstrate how, through interaction in the multidimensional problem space of robotics, the group was able to engage in problem decomposition and create a solution to one element of the robotics challenge. This small section of our microgenetic case study provides video and audio evidence of the group's learning process and development of knowledge over the course of the afternoon.
Related Content for Girls Connect: Computational Thinking & Doing with Robotics
-
2015Microgenetic Learning Analytics
Florence Sullivan
-
2016Robotics and Girls' Internalized Stereotypes
Florence Sullivan
-
2018A Focus on Girls' Math Identity for Brighter Futures
Merle Froschl
-
2021GaleForce: Girls Immersed in Robotics Learning Simulations
Beryl Hoffman
-
2019Stereotypes about Brilliance Are a Barrier to STEM Diversity
Andrei Cimpian
-
2020Theory-based Computational Analysis of Classroom Video
Christina Krist
-
2017Instructional Strategies for Engaging Girls in STEM
Vickei Hrdina
-
2017SciGirls Profiles: Real Women, Real Jobs
Rita Karl
Florence Sullivan
Associate Professor and Chair
We argue that the hands-on nature of robotics is key to student engagement and knowledge building in the area of computational thinking. Therefore, we argue that robotics (and any other external computational media – arduinos, conductive thread, etc.) are more robust learning activities for children to engage in to develop CT and their knowledge of computer science. We think it is more effective than all of the screen-based, block-based approaches to teaching students. We are basing our arguments both on Papert’s (1980) theory of constructionism, and on Bruner’s (1973) representational view of learning, which begins with the enactive representation.
What do you think? Do you agree, or disagree? We would love to debate the merits of screen only vs. manipulative + screen learning for children. Especially since the manipulatives are so expensive as to be out of reach for lots of folks.
florence
Scot Osterweil
Research Scientist
You do a very nice job of laying out the rationale for your study. It does leave me with two questions that I'd be interested to explore further:
1. Can you say something more concrete about the practices of the students that brought them to their solutions. Were they conscious of their testing and design strategies. Could they replicate or explain the process (i.e. were they thinking about their thinking).
2. While I'm inclined to believe in the power of embodied learning, I'd love to know more about why you think the physical objects was more effective than an on-screen simulations would be. I can imagine a real object might seem more "relevant," or there might be physical properties that are easier to key in on in a real object as opposed to a virtual one, but insofar as you've done the study, I'd love to know more about what you think.
Florence Sullivan
Associate Professor and Chair
Hi Scot - Thanks for your questions. I'll do my best to answer them in turn.
1) Yes, the students did think about their thinking in that they tried various strategies for locating the best spot to initialize the robot. In addition to the strategies included here, they also drew a map, and took an overhead photo. They were deeply engaged with all aspects of the problem - but, this was the one element of their problem solving that I could clearly demonstrate in a 3-minute video. I do believe the girls in this video could discuss the strategies they used, why they used them, and which ones worked best.
2) In terms of why I think the manipulatives are better, I agree with both of the colleagues who've commented below, (Joseph and Debra) that the hands-on aspect adds both interest and complexity to the activity. I also think the embodied aspect opens up ways of knowing that are not available to the students if they do not already have fairly sophisticated mental models, or mental modeling capabilities. I do think the goal of making the real robot move causes students to use their own bodies in reasoning about the problem. I like to think of Heidegger's notion of "ready to hand" and Strauss' notion of bricolage in this respect. The idea that we (as humans) use whatever is ready to hand to accomplish our goals (hence we might re-purpose and heavy object to serve as a hammer, in a given situation). This is also at the root of the idea of bricolage - re-purposing tools, a creative act born of necessity - arguably a widespread human approach to solving problems. With a motorized manipulative moving about the floor, children begin to reason about programming by using their own bodies as a stand in for the robot. In this instance, their own body is "ready to hand" and re-purposed for thinking about the problem at hand. When the child enacts the proposed movement of the robot, they can now reason concretely, an intermediary step that allows them to begin to build a mental model that will later become an abstract representation that they can draw on (without the need for embodiment). In the years that I have worked with children learning robotics I have frequently observed the spontaneous, body-based reasoning as students puzzle through a problem. However, in the two years that I closely worked with students building Scratch programs, I have not seen a similar approach. The body does not enter into it in the same way.
Well, that was a bit long-winded. But, that is how I think about it.
Joseph Gillen
" if they do not already have fairly sophisticated mental models, or mental modeling capabilities"
This, to me, is the embodiment of the argument for manipulatives. Robotics is the draw and the tool. I often use the phrase "come for the robots, stay for the science". Our program often draws children that aren't normally inclined for "traditional" science and therefore don't have a developed/internalized method for identifying a problem and reasoning through potential solutions. They come because these are fun and cool and they develop these processed without intentionally doing so.
I don't know if this is a better approach or not for those who already have a framework and discipline to learn in any given structure and I would love to find the answer to that question.
Joseph Gillen
As a veteran coach of FIRST Lego League (2011's Food Factor was actually my first year) I can wholeheartedly agree with saying manipulative+screen learning is infinitely more effective. Thinking even beyond robots in this program, the research portion requires developing an innovative solution to a real world problem--transporting and storing food, for example in Food Factor. In my experience, teams are more successful when they develop a physical object or prototype of their solution. Being able to see and touch an object is more inspiring and creates a lasting "bond" with the project that keeps drawing interest back to the project.
Florence Sullivan
Associate Professor and Chair
Thanks for your comments, Joseph. I completely agree.
Debra Bernstein
Great video, Florence. Thanks for sharing! I agree with Joseph that there is definitely value in working with robotic manipulatives. Another reason, perhaps, is that physical robots make the design space more complex by engaging students in engineering design practices around building/debugging the physical robot.
Joseph Gillen
Florence Sullivan
Associate Professor and Chair
Yes, I absolutely agree, Debra. Please see some of my other responses (above and below) for more of my thoughts on this.
Jessica Hammer
Assistant Professor
It might be possible to break down the question of manipulatives even further - e.g. physical manipulatives that are controlled by an instructor or peer (a "wizard of oz" style solution) versus physical manipulatives that are autonomous (such as robots). Would students bond equally with non-autonomous manipulatives controlled by a teammate (or even an opponent)? Really interesting question - like Scot, I'd really love to hear more about why you think the physical components are so important.
Joseph Gillen
Florence Sullivan
Associate Professor and Chair
Hi Jessica - I'm not sure what you mean by bonding with the robot, are you talking about having an emotional investment in something one has created? If so, I do think that the act of creating and controlling a robot is also an important piece of why the manipulative matters (see the more drawn out explanation above, in response to Scot).
There is something to be said about the materiality of the robotic device - a program running on a screen is very cool - but, if the thing that you program moves about and interacts with the environment, that seems to me to be a qualitatively different experience. I do think that students would also respond to non-autonomous devices controlled by others - I think that it is not just the possibility for control that people respond to, but also the sheer real-ness of the thing. The robot is a true "other" in this formulation - something to be regarded and contended with - it is different than a program that can only run on a screen.
Joseph Gillen
Robert Zisk
Graduate Student
The students seemed very engaged in the problems and developing the robot. I am curious about the learning curve with the robot and the programming. The students seemed really proficient at working with the robot, and I believe you mentioned its only a one day workshop. Did the have experience with the programming before?
Florence Sullivan
Associate Professor and Chair
This was their first exposure to robotics. The programming was not sophisticated. It consisted of forward/backward and left/right turns. No sensors were used, though students talked about the possibility of using sensors. Also, none of the students looped their programs, which in some cases, would have been appropriate - instead that just chained together the same commands. So, the programming was simple, though it looks impressive. That is one of the achievements of robots (and Scratch and other block-based programming environments), there is a pretty low bar to beginning program.
Thanks, for your interest in the work.
Deborah Fields
Can't wait to read the papers that come out of this. We are using some of your earlier research on problem-solving with robotics as inspiration for students debugging e-textiles (though we are beginning to move past an introductory level).
We're particularly intrigued by the spaces students are moving between (the physical artifact, circuit diagram, screen-code) as well as how the physicality of the code affects problem-solving. Please email Yasmin and me your papers as they are written!
Florence Sullivan
Associate Professor and Chair
Sure, Deborah, will do. We have a lot of data, so it will be awhile before work is out (probably next year).
f.
Deborah Fields
Further posting is closed as the event has ended.