NSF Awards: 1325004
2016 (see original presentation & discussion)
Informal / multi-age
Educators working in informal learning spaces must attend to learning goals without the structures provided by a formal learning institution. While the need for STEM learning opportunities in informal environments is increasingly apparent, educators and mentors in these spaces often do not have STEM content specialization or expertise. To address these challenges, as part of work led by PIs Nichole Pinkard and Denise Nacu, we are co-designing data visualization dashboards with mentors in an informal, blended learning environment.
The Digital Youth Divas (DYD) program is designed to engage girls in urban Chicago, especially those from underserved communities, with computational circuitry and programming through fabrication and design. DYD, run by the Digital Youth Network, incorporates project-based learning and an online platform. Five program mentors engage with the girls in the face-to-face classroom each week and two mentors are responsible for online assessment of computational projects submitted to the online space.
In this video we share educator data dashboards that pull live data from the online DYD platform, displaying metrics to help mentors manage their blended-learning spaces and support their young STEM learners. We focus on sociotechnical dimensions, including platform features and technical capabilities that support mentorship in informal STEM communities alongside authentic DYD mentors practices, intentions, and challenges. Together, researchers, designers, and practitioners are determining which metrics are useful in this context, how information should be displayed in order to be most effective, and how mentors use the data to support individual girls in their STEM learning trajectories, and build community engagement and success across the cohort.
Caitlin K. Martin
Senior Researcher
This is both a story about making data visualizations accessible and actionable for educators in a way that makes sense for their own unique goals and community, but also about the importance of collaborative design.
Along with increasing expectations that educators use digital curriculum and online resources, are implications about how resulting learner-generated data should be utilized. What related successes, opportunities, and challenges have you faced? We would love to know.
Have you been asked to look at and make use of online data? What are the disconnects between what the data shows you and what you want to know? Do data and visualizations help to level the playing field, identifying and supporting underserved learners, or are these systems not even designed with such learners/environments in mind?
Brian Drayton
First of all, it looks like people are having fun and working hard!
As I watched, however, I thought about how I might ‘inhabit" the role of mentor. It looks like the student-to-mentor ratio is fairly high, and it was intriguing to see the mentor sitting at the computer while all that activity was happening in front of her. Is the system a way for the mentor to enter notes and track what’s happening (so, a project-tracker)? But your abstract talks about learner-generated data. How are the data collected, and what more do the mentors get from it that they would not get by floating and interacting with their teams of girls, and making notes on a clipboard-based protocol?
Caitlin K. Martin
Senior Researcher
Hi, Brian. Thanks for your questions. The implementation in the video is a Saturday program with 5 classrooms each with one mentor, one jr. mentor, and about 20 girls (so 10-to-1 student-mentor ratio).
In this space, while the role of the mentor is absolutely first and foremost face-to-face, as they build community, offer individual encouragement, and help the girls succeed in the project work, there is also interesting data captured as girls navigate the program online platform. In this platform, girls are presented with scaffolded project challenges, are given learning resources like how-to guides, can ask questions of mentors outside of class time, and can view and comment on the work of others. Girls’ activity online is logged on the system.
While mentors easily navigate the just-in-time help needed in the moment, they wanted a larger overview understanding of where the different girls in the program were on the larger learning pathways (or sets of activities). The dashboard views allowed them to look outside of class time at where the individual girls were on the larger program learning pathway, and reflect on who may need additional help or where people could advance if they were done before others.
Pati Ruiz
Dean of Studies
Presenting information to mentors so that they can support individual learners is essential. It is great that you are able to get feedback from mentors and have that drive the improvement of the tool that you are creating. I have some of the same questions that Brian mentioned above, how is the data getting into the tracker?
Also, is this a tool only for the mentor? Is it (could it also be) a metacognitive tool for the learner? Will the learner see their progress over time and use this information as part of their learning journey?
Caitlin K. Martin
Senior Researcher
Hi, Pati. Thanks for your clarifying questions — 3 minutes is a fun way to get the story out there, but it was hard to fit a lot of specific information in! This is part of a larger study looking at educator dashboards in different types of learning environments.
In the environment showcased in the video, the online platform tracks and logs specific user actions on the system in a system database. We have linked that database with Tableau, a data visualization and analysis tool that we use to create flexible and dynamic and evolving dashboards for the educators to look at student activity in the online platform.
The girls have their own view of where they are in the pathway on an individual level through their profile page on the system, as well as building a portfolio of work as the program unfolds.
Pati Ruiz
Dean of Studies
Ah, this makes sense. Thank you for this explanation – very helpful!
Jenna Marks
Doctoral Student in Cognitive Studies in Education
Hi, Caitlin. I’ve worked a bit on building teacher dashboards for blended learning programs and also am currently in the midst of a design process building an exploratory learning environment through a fairly collaborative design, so I figured I might try to address some of the questions you posed.
I agree with your mentor that it is SO CRUCIALLY important to get feedback from your users on what they want and need from a data dashboard. You might find quite varied use or that a piece you thought vital is never used by your teachers/mentors. At this point, you can decide that 1) the piece is unimportant and should be removed, or 2) this piece is important and we must train teachers on how to use it. I believe that including some training and guides alongside dashboards are critical in helping a teacher to interpret the data story she is viewing, and in my own work I’ve done so by creating a variety of student types and illustrating how I might understand their progress through the dashboard and data.
In terms of disconnects, I think it depends on your data and what you are aiming to learn. Of course we cannot reduce a student to a few bar charts. One common issue I’ve seen is the lack of longitudinal information. Unless I know each student well, it often helps to see the pattern in their behaviors and scores over time.
I’d love to hear anyone’s thoughts on any of the above as well!
Caitlin K. Martin
Senior Researcher
Thanks, Jenna, for your thoughtful reply! Yes, we are really interested in longitudinal data as well, especially as these sorts of learning communities can really grown and change over time, reflecting individual changes sometimes for both adult educators and youth.
Denise Nacu
Faculty
Thank you for your response, Jenna! Adding to Caitlin’s comment, we are also interested in bringing varied sorts of data into the dashboards to provide a richer picture of the individual in order to aid mentors who might have limited prior interactions with youth. For example, we are exploring ways to bring together attendance information and survey responses about youth interest and experience in a way that is useful for mentors.
Avron Barr
Consultant
Thanks for putting together such a nice video, Caitlin. Besides the dashboard “snapshot”, have you thought of any other management tools for mentors and learners that might be extracted from the data you collect?
Caitlin K. Martin
Senior Researcher
Hi, Avron. Thanks! Yes, we are thinking about data visualization and dashboards across various types of learning environments — what sense can be made of click data in ways that educators (and learners) find useful and actionable. So not just for management (which this video does highlight and which is more traditional in terms of the educator role in such an environment) but also for insights in terms of activity and opportunities that could be considered related to 21st century learning, like interaction and collaboration, utilizing resources, and persisting in challenges. In other words, visualizations that might spark some new ideas about the types of activities learners are engaging in online, and helping to leverage these sorts of cross-curricular practices that have been identified for STEM careers and tie them to ideas about what next (brokering new opportunities based on prior activity/interests/etc).
Roxana Hadad
What a great project! Is there time set aside to help the students understand the data presented or learn more about data visualization? I realize this isn’t the focus of the project, but thought there might be interesting for the students to both learn about data visualization and learn about themselves as learners.
Caitlin K. Martin
Senior Researcher
Such a good question! Right now, our focus for the dashboards in on educators. The platform itself does have a progress visualization that the girls use to see where they are along a learning pathway, which they are encouraged to use to monitor their own progress and get help as needed.
For the educator dashboards, we are utilizing a little more data, and trying to envision some views that are a little farther out there than just classroom management (see the response above). Even with adults, there is a learning curve to interpret visualizations in ways that make sense to them. For many educators, this is not something they were trained in although they may be asked to do this sort of interpretation using data from school or district systems. As part of our work we are developing PD materials to prompt reflection and discussion around data, looking to support educators to really select platforms that collect the type of data that they are interested in and use that data to better serve their community of learners.
Leslie Herrenkohl
Hi Caitlin I loved learning about the program and the digital dashboards! A similar challenge of understanding youth learning and using this kind of data to support mentors to reflect on facilitation emerged in our project so it is fantastic to see others thinking about this issue! (see our video here: http://stemforall2016.videohall.com/presentatio...
Thanks for answers to earlier questions that helped me understand more about the system. I love that you are linking with tableau for the visualization part. What do you track in these displays? Are you using existing database tools to create the data to link with Tableau? Great work!
Further posting is closed as the event has ended.