Community Support for Constructionist Learning
Amy Bruckman
This journal paper is an examination of the MOOSE Crossing system designed by Amy Bruckman and a host of people at MIT’s Media Lab. Based on MOO, a MUD type environment developed for collaborating researchers at Xerox PARC, MOOSE Crossing is designed as an environment in which young children can create worlds and object that exist in them by learning to program in a basic scripting type language. Children logged in and participated from after school programs around the country, at the MIT Media Lab, and at home. This paper focuses less on the technical aspects and is written more as a “lessons learned” summary of some of the findings in the approximately 3 years the program ran.
The basic premise is that people benefit from MOOSE because it is a constructionist environment in which children, and occasionally adults, work together to construct a representation of themselves by writing scripts that describe the locations and objects in their world. There are 5 important aspects why Bruckman says that people use MOOSE: role-models, situated ubiquitous project models, emotional to overcome technophobia, technical support, and an appreciative audience for their work.
The key is that people play for mainly social reasons; it’s only important to make an interesting room for yourself so that you can show it to all of your friends. People enjoy being learning from more experienced members as well as teaching new members. Generally, after a help session, both the tutor and the pupil would hang around and talk for a while, both having made a new friend. Also, the interface hides certain features, such as age, gender, social status, etc., that might otherwise limit interactions. In building this constructionist environment, “creating natural opportunities for casual, social interaction is of central importance” (Bruckman, p. 84).
There are a couple of ramifications to the EcoRaft project. First, Bruckman found that collocated activity worked to complement individual activity for those who connected from home. This has implications for the game deployment. Also, while not totally open-ended, our participants should be constructing a biologically diverse rainforest together, and that act of construction should be a trigger to assist in the learning of ecological concepts. Perhaps giving the participants some more control in how to develop the rainforest as they restore it, such where to plant trees, etc. While participants currently are able to choose what species to bring to an island, we are reinforcing bringing as many species as possible, which doesn’t leave them much room to choose some and not others. Perhaps the later possibility would work better in a system where we had multiple shade giving legumes, multiple pollinators for those legumes, etc.
As far as methods, there's not much to be taken here. Basically, she logged all the text ever typed on the system and interviewed the participants. However, it also seems that she engaged with participants in the system, which may be something significant we may want to do, however this is a rather different application and that approach may not work for us.
The effect of tangible interfaces on children’s collaborative behavior
Authors: Danaë Stanton, Victor Bayon, Camilla Abnett, Sue Cobb and Claire O’Malley
link Abstract: The physical nature of the classroom means that children are continually divided into small groups. The present study examined collaboration on a story creation task using technologies believed to encourage and support collaborative behaviour. Four children used tangible technologies over three sessions. The technology consisted of a large visual display in which they could input content
(using Personal Digital Assistants (Pda) and a scanner), record sounds (using RF-ID tags) and navigate around the environment using an arrangement of sensors called ‘the
magic carpet’. The children could then retell their story using bar-coded images and sounds. The three sessions were video recorded and analysed. Results indicate the
importance of immediate feedback and visibility of action for effective collaboration to take place.
***
The KidStory project was designed to examine how children interact with each other when working in small groups using tangible technology. Throughout the project, instant feedback was incorporated so the children were always aware of the consequences of their actions. The software, KidPad, is a collaborative authoring tool designed specifically for children. It allows the subjects to create stories using links and also to input images and sounds to their stories. PDAs and scanners were used to input pictures and microphones were used to record a sound. Additionally, an
RF-Tag was placed on the tag reader while the sound was being recorded. When the RF-Tag was removed, a barcode was printed out on a label for later playback. The subjects used a "magic carpet" (12 floor sensors placed in a square with a rubber mat covering them) to navigate through the game. Whenever a subject interacted with a system, be it by recording a sound, scanning an image, or using a PDA, an icon for each technology was displayed on the screen.
Researchers observed three types of collaboration among the children: unspoken collaboration, aiding & tutoring one another, and tactic development-problem solving. The children collborated without verbal communication to help one another use the technology. For instance, problems with the scanner were avoided because one subject would anticipate it before any action was taken. The researchers believe that this lack of verbal communication indicates that "the design of the technology encouraged collaborative behavior." Similarly, the subjects spoke to one another to help with the uses of the technologies available. Oftentimes, the researchers observed one child taking the role of the tutor to help his or her classmates. The researchers also observed how the children, by trial and error, managed their time and resources to use the technologies. For instance, when scanning a picture, they realized that there was no space available on the screen for images and thus quickly moved the screen to create more space. By trial and error, they discovered that they could use the time delay of the scanner to make more space available on the screen.
***
It is important to note that it is not specifically stated whether or not the researchers used any written or verbal evaluation methods. It is known, however, that the researchers observed the subjects and their interaction with the project. I think that the observation aspect of this experiment can be useful for our project if we let our subjects interact with the exhibit without any instructions. However, I think our evaluation method would not be complete without some verbal/written communication with the subjects.
Pianos Not Stereos: Creating Computational Construction Kits
Resnick, Mitchel
Bruckman, Amy
Martin, Fred
MIT Media Lab
This paper presents a few interesting and compelling theories about education and how those theories have been used to inform the design of a number of educational tools. The premise of the paper is, which would you rather your child learn to play, the piano or the stereo? The stereo is easy and gets immediate response, but the piano has more expressive range, is more personal, and ultimately more rewarding. The authors then extend this metaphor to computers, arguing, using Papert’s constructionist approach, that we should create computational construction kits that give kids the ability to create things with the computer. They present idea that construction can provide two kids of connections, and construction kits that allow both kinds are more beneficial. Personal connections link activities with users’ previous experiences, making them both more approachable since somewhat familiar and more meaningful. Epistemological connections link together domains of knowledge and encourage new ways of thinking through these connections. The authors present three examples that use these principles. Rather than give specific, low-level details, they describe the basics of each project and how it adheres to these principles.
The first project, Programmable Brick, allows kids to program behaviors into lego models. There is a personal connection, because legos, and toys in general, are part of a child’s everyday culture. There is an epistemological connection because, as the children build the devices, they start to wonder about the differences between animals and machines. For example, one child built a machine that is attracted to light. Using two photo sensors as “eyes,” if it receives more light on the right, it veers to the right, and similarly for the left, much like Braitenberg’s machines. Such feedback mechanisms are usually not taught until university level courses, but by allowing the child to make his own epistemological connections, he was able to figure out the feedback structure on his own.
The second project, StarLogo, deals with developing complex behavior using decentralized models. Rather than developing a single, overarching control structure, control is distributed among many individual elements, as with ant colonies, traffic conditions, market economies, etc. This environment makes a personal connection, because everyone is familiar with such systems, although they may not be familiar with how they work. It also provides epistemological connections by encouraging different types of thinking about behavior. For example, two high school students who developed a traffic simulation discovered that traffic jams do not always come from a centralized cause, such as an accident or broken bridge, but often emerge as the result of distributed behavior. Traditionally, such complex systems are described using differential equations, which are very abstract and deal in aggregate quantities rather than dealing with the individuals. This project makes these complex systems approachable and comprehensible to young children.
The third project, MOOSE crossing, features an online, MUD-like world where children create their own personas and locations. It features basic scripting, so that children have to learn to program in order to describe themselves and the places where they are in the virtual world. This project creates a personal connection because it allows children to describe themselves in terms that are meaningful to them; the environments created ranged from starships to islands resorts. The project also creates epistemological connections, because it integrates computer programming and creative writing. One girl, who reported that she did not like math or other analytical tasks enjoyed programming, because she saw it as just a different type of writing.
One of the key properties of these projects is that they allow for emergent experiences. No matter how much the designer tries, the user will always try to do something that was either not intended or not expected. Environments that allow for a wide range of activities, rather than restricting to a few pre-scripted ones, so that users can construct their own experiences.
I'm not certain how much of this can or does pertain to the EcoRaft project. We should definitely cite some of their ideas about constructionism, but we're not letting the users create anything from scratch or nearly as open-endedly as Papert's ideas would suggest. Maybe we want some way for kids to place seeds, within certain restrictions, so that they feel more ownership in having helped design the look of the forest. Ultimately, these are some cool ideas, but I'm not sure we'll incorporate them much in this project.
Combining Handhelds with a Whole-Class Display to Support the Learning of Scientific Control
Authors
Tom Moher University of Illinois at Chicago, Chicago, IL
Xun Ding University of Illinois at Chicago, Chicago, IL
Jennifer Wiley University of Illinois at Chicago, Chicago, IL
Syeda Hussain University of Illinois at Chicago, Chicago, IL
Preeti Singh University of Illinois at Chicago, Chicago, IL
Vasisht Srinivasan University of Illinois at Chicago, Chicago, IL
Diane Conmy Abraham Lincoln Elementary School, Oak Park, IL
http://portal.acm.org/citation.cfm?id=766048 Abstract: Third grade students used wireless handhelds and a large shared display to discover strategies for control of variables in scientific experiments. The technology suite supported activity requirements including synchronous individual control, face-to-face discourse, and instantaneous display updates. In an empirical study, students demonstrated learning in both original and transfer domains.
***
In this study, the researchers allowed their subjects to navigate through a large shared display with networked wireless handhelds to help students learn about a fundamental concept of experimental science, the control of variables (COV) strategy, and hopefully adopt this strategy easily and successfully. The researachers developed an activity called "Who's Who?" in which each student simultaneously controlled an independent variable on a projected screen of a grid consisting of "pixels." Each pixel on the grid is controlled by its own handheld PC. When a button on the handheld is pressed, the corresponding pixel changes colors. The goal for the project was to ask the group to modify the color of the pixels until it reached a final configuration state.
Before the experiment began, the subjects, a group of third-grade students, were given a written test so researchers could evaluate their existing knowledge of the COV strategy. The researchers conducted two experiments on consecutive days after the written test, with students mapped to different pixels for each experiment. After both experiments, the researchers conducted structured interviews (pre-determined questions, methods, etc.) with the subjects to assess whether or not they understood what had happened. These interviews were videotaped.
The results of the two activities were very dramatic. For the first experiment, the subjects took a little more than half an hour to complete the task. For the second experiment, the subjects took less than 4 minutes.
The researchers believe that most of the task strategy learning occurred during the first session. This is obvious since the little time was passed before the subjects completed the activity the second time. The subjects articulated a "one at a time" algorithm after each activity and their mechanism understanding increased more than 20% between the two activities.
***
I think the written test was a good way to evaluate the subjects' knowledge on a specific topic. I don't think it would be very effective for the EcoRaft Project, however, because our questions are open-ended and it would probably be more successful to conduct a semi-structured interview. However, I do think it would be important to review the subjects' improvement in learning strategies, like mechanism learning, conceptual learning, etc.
The Round Earth Project - Collaborative VR for Conceptual Learning
Johnson, A., Moher, T., Ohlsson, S., Gillingham, M., The Round Earth Project: Collaborative VR for Conceptual Learning. In IEEE Computer Graphics and Applications, vol 19 no 6, November / December, 1999, pp. 60-69.
http://www.evl.uic.edu/aej/papers/cga99/cga99rev.html"The Round Earth Project is a collaboration among researchers in computer science, education and psychology. It investigates two alternative pedagogical strategies for teaching children that the Earth is spherical, and the implications of that fact. One strategy, which we term the ‘transformationalist’ approach, attempts to effect conceptual change by breaking down the children’s prior models. The alternative ‘selectionist’ strategy, in contrast, attempts to effect learning in an alternative setting (in our case, a small diameter asteroid), free of pre-existing biases, and to relate that learning back to the target domain: the Earth."
***
Virtual reality is used to simulate much of the action in the project. The researchers hope that virtual reality will affect the children's capacity for conceptual learning. They also hope that introducing new, fundamental ideas will replace the learners' existing ideas. Three pilot studies were executed. The first study was to familiarize the subjects with the exhibits. For the second study, the researchers conducted pre-test interviews to assess the subjects' knowledge of the topic at hand and then questioned the subjects after the interaction. The results from the second study were "disheartening" because the subjects had not yet abandoned their existing ideas. In the third study, the researchers brought in 2d models to demonstrate the difference between 2d and 3d models, but the subjects did not maintain consistency between the two models. Although some of the subjects still held on to their existing ideas, others demonstrated temporary effective learning immediately after the experiment but not the next day and others appeared to reflect persistent learning of some of the target knowledge components. Each case study was designed to probe for understanding of particular topics. The researchers designed a set of questions to ask before and after the subjects' interaction and measured whether or not they understood the topics at hand. In particular, the researchers asked them whether or not they understood the concepts behind particular words such as "horizon" (which meant little to the subjects). The researchers created both written and verbal questions to test what the subjects learned. The subjects treated the experience more like a game and this negatively affected their ability to replace their existing ideas with new ones. However, for the subjects who succeeded in changing their ideas did so because of the changes the researchers made in their protocol.
***
In this study, the researchers continually improved their evaluation methods. Their ways of evaluating included familiarizing the subjects with the exhibit [by interaction and also by providing supervised guidance], questioning the subjects on what happened and how it came to do so, introducing new methods when the previous studies failed to show any improvement [i.e., bringing in 2d models to distinguish it from 3d models], and post-game interviews.
In terms of the EcoRaft Project, I think it might be helpful for us to introduce some sort of guidance during our subjects' interaction with the installation if need be. This can be done by providing sparse hints throughout the gameplay. However, the goal for the Round Earth Project was to dispel any ideas that the Earth is flat-shaped and not spherical. Our goal is not to dispel any ideas but to encourage a more ecological way of thinking.
