I think you are right, Dennis - there is no doubt that eventually, we would need some probabilistic reasoning technique to do any interpretation of the data we can get, and a Bayesian approach seems like a good fit. At this stage in the project, we aren't quite to this point in yet since we are currently capturing a lot of video of kids interacting with the exhibits and analyzing it. We need to find out which movements seem to best reflect learning at each site, code it properly, and then (hopefully) we could identify ways to implement a probabilistic classifier based on that labeled data.

And what a great thought re: stealth assessment - we hadn't considered that idea in this context yet, so thank you! I am a huge fan of Val Shute's work (my students read several of her papers each semester).... 

Hi Jamie

 We are just piloting an observation tool now and would love to learn more about Bayesian networks. Thanks Dennis for the suggestion.

Hi Rebecca,

Your question is spot on - something we have struggled with and spent much time discussing /reading about. Below we share our latest thinking but very much welcome your thoughts.

Here is our (very much) working definition: we consider learning as changes in embodied interaction demonstrating an improved ability to predict the world and communicate predictions with others. Some thinking below (decided not to cite throughout but very happy to share our influences if helpful!).

We found it helpful to start by defining how learning is manifest; how we can capture learning. For this, we consider a change in children’s embodied interaction, adopting a multimodal perspective. So, for example, learning may be a change in how children manipulate objects, talk, gesture, or perhaps a simple change in eye attention. These changes may be small but significant, hence often the need for close inspection. (e.g. video). Whilst much science research has focused on speech, we hope to add to exciting work looking at other modes, significantly children’s gestures which can represent certain scientific relationships. This is particularly valuable for our population (young children) who may be able to express themselves through actions and gestures before they can through speech, although work shows the relevance for older learners.

Our next challenge is a bit more tricky – what types of changes in embodied interaction do we consider learning – how can we agree on what changes we wish to encourage? Here we wanted to avoid circular definitions by using more abstract terms that themselves need unpacking (e.g. concept; understanding) and have been influenced by work on predictive processing and social constructionism. We propose that learning is improved ability to predict the world and communicate predictions with others.

The first aspect tries to capture diverse work such as predictive modelling or perceptuo-motor integration. Such learning may or may not be explicit. The second part of our working definition emphasises the importance of being able to communicate knowledge with others. To communicate our predictive models of the world, we need to develop ways to describe, explain and make predictions (explicitly).

Therefore, we are investigating children’s learning by examining changes in embodied interaction that evidence an improved ability to predict the world and communicate predictions with others. Examples might include changes in eye gaze showing children are improving their predictions of the movement of an object, or changes in gesture showing children are improving their ability to communicate a scientific relationship. Changes may also include children appearing more interested (i.e. greater visual attention) in certain patterns; or asking relevant questions (i.e. improved communication).

We are continuing to develop our thinking in this, in particular, where emotional aspects sit in our definition. Currently, we see emotion aspects as fundamental – they shape if and how children interact, or communicate; however, we feel that emotions are core to the conditions for learning, rather than integral to our definition.

As said at the start, our thinking is still very much ongoing, where we really do look forward to hearing thoughts on how others have addressed this challenge of trying to define learning from an embodied perspective. Thanks again for your question!

Just chiming in to say I'm very interested in Lesley's question, as well! You mentioned in the video that you are looking for cross-cultural signifiers, and I'm interested to know more about how those are emerging (or not).

Hi Lesley, Kalie, and Victoria. Thanks for your questions. We are working across multiple museums in both the US and the UK, with each museum serving a number of communities. While we might be able to compare US to UK contexts, we are also looking at differences in the communities each museum serves. For example, as Victoria hinted at, Frost Science uses interpretive labels in English and Spanish and serves the diverse communities in and around Miami. That said, we are in the early days of collecting data, so we don’t have any firm answers yet. But our research design attempts to acknowledge that there is diversity in how adults from different cultural groups facilitate their child’s learning in a museum setting. Do you have any examples from your work that might apply? 

The project also provides an exciting opportunity to focus on other modes of communication, like gesture, and explore how this may benefit the communication of science ideas across cultures/ language. If young children are able to demonstrate the ability to abstract key science ideas through specific gestures (say, rubbing hands together to illustrate ideas associated with friction), there are interesting questions around cross cultural use of gesture and their potential for sharing/ communicating around science.

Hi Jessica,

Many thanks for your comment. Touch screen devices like iPads or smart boards can offer a lot of interaction - sounds and swiping movements (some designs even respond to moving the device around) - and there are various projects looking to create new screen resources drawing on embodied learning. That said, our project is investigating whether there are additional benefits from more expansive, three dimensional, movements that can be achieved through physical actions and gestures. Museums / science centres often have the advantage of wider space and larger budget to create these whole body interactive experiences, although teachers have been using physical materials (e.g. balance scales) for decades. Where I think there are really interesting developments:  a) increasing range of physical digital products being created for classrooms and b) research showing the potential impact of teachers' gestures. Personally, I think the latter is most exciting as it is a) free b) can be done anywhere c) something nearly all teachers do anyway. One shared research goal is to develop guidelines for what types of gestures may be most beneficial. Our project hopes to contribute to this understanding, and whilst in the science museum context, we hope will have wider implications, notably for classroom teachers.

Hi Andrew,

I think an important part of the question from Jessica is also about the type of learners. I got the same question a year ago when I gave a presentation on the topic at a festival. I am very curious if you have found something about children that would still prefer to read for instance. Is it something really universal or do we need to tailor our classrooms more in the future, some like to learn embodied while others will be doing more traditional learning? Probably not easy to answer, but curious to see how one can even start looking for an answer to such a question.

BTW really really nice project, hope to (keep on) work(ing) on similar topics in the future, and come up with even more exiting findings.

Ah, sorry I missed this, thanks for flagging this Robby. It is early days in empirical work but there do seem some interesting differences in how children engage with physical exhibits, and how they communicate through words and gestures. I do, however, share your thinking in terms of the difficulty of answering how we respond to differences as educators - do we adapt to children's preferences (e.g. reading) or encourage variation? There is also the danger of repeating past mistakes of labelling children, e.g. 'a kinaesthetic learner'. 

Where we think our project can contribute is through greater recognition of ways children can communicate their thinking through interaction and gesture, and how we might develop experiences accordingly. As we work with more children, it will be great to identify patterns of differences between children (and their supporting families attending museums) that, as Jessica indicated, would be valuable knowledge for educators.

Thought I'd add a bit to this discussion - the question of individual differences is always interesting to me in learning sciences research. For example, if you look at the 3 children being interviewed at about 1:45 in the video, the two on the left are really expressive - they observe and model each other, almost even compete to share their ideas by standing. It's wonderful. The child on the right, however, seems largely unaffected by any of it - she is listening (I think) but not doing any meaningful gestures or movement (that I can detect). This could be a personality difference (she might be shy), or perhaps she is more careful and would prefer to hold back until she can confidently talk about the exhibit she was in. It's really interesting to think about and as psychologists, try to understand. Ultimately, our project is about (1) how do children move their bodies when learning science and how does it relate to the underlying concepts, and (2) how can we promote movement that can reinforce those ideas/concepts. In our pilot data we are seeing this range of movement, both in terms of how much they move, how they respond to designs or facilitators that encourage movement, and what sorts of contexts allow them feel comfortable enough to express their ideas with movement (e.g., props in interviews often make a difference).  

Hi Jamie,

Great questions and certainly ones we hope to better answer toward the end of the project. The short answer to your first question is yes, we are very much thinking (and learning we hope) about differences between types of experiences. If I share some initial thoughts. Embodied learning suggests that sensory and action experiences can help develop cognitive resources we can draw upon in our thinking, reasoning and communicating science ideas. Hence most activities, from just talking, to looking at pictures, can provide such experiences. Our work is looking at the additional value of moving your body in particular ways when learning. This is much more than just being more engaging for children, for example, we know that both children and adults often express particular movements in their gestures when asked to explain their thinking. Both physical exhibits and virtual ones can encourage children to move their bodies in certain ways, where each has certain affordances and limitations. For example, gesture-recognition exhibits can respond and hence encourage very particular actions. But they are limited in providing tactile feedback of the result of that action. Screen animations can provide a very rich dynamic visual image, but responses are programmed rather than being caused directly by user actions. So, I guess this leads to the slightly familiar cop-out research response that it depends. Some of our exhibits benefit greatly from rich dynamic imagery - e.g. moving animals; others from simply physical cause and effect, e.g. balance board. It is exciting for us to be exploring some designs that try to leverage both forms - adding a virtual layer to hands-on exhibits. 

Regarding your second question, we are still developing our coding of gestures, but it likely we would code the types of gestures you refer, because a gesture that appears playful or poetic would likely be meaningful for children, and this is fundamental. That said, I guess we will be wanting to understand how such gestures related to more formal or shared, gestures that appear to represent concepts. Thinking of examples, we would be interested in gestures where children are drawing another's attention to something they find interesting, or gestures that show how children are relating previous experiences to their exhibit experiences. Great to hear if you have any examples of these types of gestures in your context.

Hi Preeti! As someone who comes from the museum world, I've learned (and am still learning) so much about embodied learning or embodied cognition. Every time I hear my fellow Co-PIs talk about it, I see it in a new light and gain new understanding. It really is about reframing how we view cognition. It's not mind or body, but mind-body as a unit. So learning itself is embodied, you learn with your body and your mind. Even when you are an adult and you've been trained through school and the modern office setting to learn (and produce) sitting down, your mind-body connection is still at play. There is evidence that your mind is processing information as if your body is moving even if you aren't moving.

I'm sure someone else could chime in about why the embodiment research privileges gesture, but my understanding is that gesture can be a "check for understanding" as Judy said above. So with this age children especially, by viewing how they gesture (or move more broadly) you might be able to get a sense for how they understand a concept that they can't put into words. As you point out, modeling can happen with tools like the foam balls and light. But if you could give someone a model that they take with them, how powerful is that! So you were on the right track with your students! 

Hi Preeti, I'll just add a bit onto Susan's excellent overview. I haven't been working on the embodied learning problem as long as other team members, like Andrew Manches and Robb Lindgren, so a lot of his has been new to me as well. I have noticed that understanding embodiment takes time, and part of that is due to societal and academic focus on information processing and cognition. We want to open it up in this project, and think about learning differently - we want to take all of the great work that has happened in the lab and apply it to museums. For example, if you are curious, Prof. Art Glenberg at ASU is on our advisory board and runs one of the most famous embodied cognition labs in the world. One of his foundational studies had children reading stories - in one group, they sat and read quietly, in another they acted out the story with toys as they read, and in a third group, they acted out the story physically (as if in a play). Well, when they tested things like comprehension and recall, the group that acted it out came out on top - the results were clear. Toys did well too, but the physical embodiment of the story gave them a concrete/tangible experience that they could associate with the story. This and other compelling examples from the lab played a big part in getting our project going.

I'll add a bit more on how I think about "movement" and "gesture". There's not going to be a single answer that will work nor is there a definition of embodied learning that I can share that is widely agreed upon. In our recent advisory board meeting, I shared two things relevant to your question. 

1. We aren't ruling out any movement in our research - whatever we determine is meaningful and relevant, we want to code and understand. This means if it is how kids run around, jump (say to model an animal on a screen), how they move their hands, etc., it is important from an embodied perspective. As Susan said, gesture (with hands) has had a very privileged role for some reason, but we are just including it as one kind of movement out of a really long list. 
2. We also want to understand how to promote those movements that turn out to beneficial in all of the contexts we are studying, so we will think about exhibit designs, simulation design, facilitator strategies, signage, and whatever else emerges as relevant. 

Hope that helps!  Please follow up if you'd like more discussion, we are loving this! 

I will let those on the team who have experience training facilitators respond, Jamie (I haven't done that myself), but will note that the adult in the video demonstrating gravity and friction (at Science Museum London) is a teacher who also uses gesture extensively in her explanations - she was so good she made it into our 3min video!  Our UK partners captured the explanations from a variety of teachers and facilitators in our first year, and learned a great deal. There is a robust literature on this in mathematics, you may be aware, showing that teachers who gesture appropriately (for things like slope, proportionality, and other things) are more effective. Our primary focus is on how pre-K learners are moving at these exhibits, but we are also very interested in how the movements and gestures of others are influencing them, and this is one major component to that thread of the work. Thanks so much for your support and for watching multiple times, we really do appreciate it!

Hi Jamie,

Would be lovely to know what type of gestures the high school explainers were using. One interesting thing we have noticed in our project that may be worth further exploration is how many of the gestures used by our participant facilitators to explain science concepts are similar to the gestures used by the signing community for science concepts, for example, https://www.signbsl.com/sign/gravity . It will be good to evaluate in our project whether gestures from the signing community offer a powerful, freely accessible resource for facilitators. 

Thanks Amanda. Being on this project did make me think of when I taught in K-2 and we were encouraged to use Makaton https://www.makaton.org as many children had speech and language difficulties. It is interesting to think how an approach to signing to support communication may be applied to STEM. 

Hi Nick,

Thanks for your comments; very much agree. Here are my personal thoughts on reasons to be concerned about the recent push for STEM; reasons to be optimistic; and how our project may contribute.

History shows a contradiction in what we say we want to promote in learning (creativity, play, self-directed inquiry, collaboration) and often what we do ("stuffing" children with particular facts and procedures). Unfortunately, much is attributable to assessment, which can drive learning. We want to capture what children learn, and it is much easier to capture retention of facts and procedures (e.g. written tests). PISA assessments not helping. The good news is there are exciting approaches to capturing children's learning, for example, capturing and storing short video clips for comparison, or seeing how children progress in a simulation. Our potential contribution is to focus on ways to capturing learning in how children interact and explain their thinking. This may be particularly valuable for children who are less 'practiced' in formal forms of assessment. 

So I am excited by the shift to include more STEM, but cautious of how we choose to assess. Creative assessment will empower creative pedagogy, ultimately allowing student to play to learn more. 

Hi Nick, thanks for your comments, I'll build a bit on Andrew's reply. There is no question to me that the societal emphasis on reading and writing instruction at early ages is supplanting free and open play time, time that critical social and emotional skill development. What is even more disheartening is that there is evidence this doesn't work - even if you start formal learning of reading and writing earlier, downstream benefits are minimal - things start to balance out later, and then you have kids with impoverished self-regulation and emotional skills. I'm making it sound dire, but there are researchers studying this carefully and they are ringing the alarm bells now. For example, Roberta Golinkoff and Kathy Hirsch-Pasek are personal heroes of mine and have written extensively about this mistake we are making as a society. I think you would love their work - it's easy to find talks on line from both researchers.

Zooming in a bit on our project, the emphasis on STEM and extent to which play is possible are somewhat orthogonal (or should be). It can certainly be both, and at the same time it doesn't have to be. In our project, we combine STEM learning with playful/social interaction, you will find links to artistic and musical aspects in the exhibits we work with, and all of the museums represented on the team provide rich environments that focus on the interconnected nature of the world. One of the main points we have in our project is that moving around is good, but paying attention to those movements and looking for the meaning they may carry is better. Your 2 young ones use their gestures and movement to communicate, perhaps even more effectively than they do with words (depending on age), so we're shining a light on that fact and hope to help with future pre-K exhibit designs.

Thanks again for your thoughts and interest!!!