The transition from a student understanding data they collect themselves to understanding someone else’s data, particularly complex and often messy professionally collected data, is an important one. The Streams of Data project, a collaboration between the American Geosciences Institute and EDC’s Oceans of Data Institute, seeks to understand how professionally collected scientific data can support elementary students’ development of data literacy skills that they will need in high school and beyond! An interdisciplinary team of educators, researchers, and scientists are conducting baseline research to understand students' natural affinities for making meaning of complex data and phenomena; develop and test scaffolding activities that minimize barriers to analytical thinking with professionally collected data; and assemble promising approaches into a multi-activity experience to examine the degree to which the combination of activities allows students to do productive work with professionally collected data.
The COVID-19 pandemic has dramatically altered the education landscape, including the ways in which we study innovative approaches to teaching and learning. This video shares our team’s efforts to adapt our original workplan to the abrupt end to work-as-usual and the myriad, ongoing uncertainties that resulted.
Science, Research / Evaluation, Addressing NGSS
Oceans of Data Inst, AGI, J. Sickler Consulting
Discovery Research PreK-12 (DRK-12)
Related Content for Streams of Data: Adapting Research During COVID-19
Discussion from the 2021 STEM For All Video Showcase (22 posts)
Ed Robeck
Lead Presenter Director of Education & Outreach
May 11, 2021 | 01:47 a.m.
Welcome to our video about Streams of Data. This project will develop and test an approach that demonstrates how fourth grade students, supported by scaffolded learning experiences, can begin developing the analytical thinking skills necessary to make meaning from complex, large-scale, professionally collected data. The work will focus on analytical thinking as a subset of data literacy, specifically understanding the nature of data, asking and answering questions using data, and communicating with and about data.
We've been challenged to keep the project on track during the COVID-19 pandemic. Our video conveys some of our reflections on those challenges and how, with support from our partnering school systems, we've been able to keep the research going.
We'd like to hear about how others have navigated similar challenges with school-based research. We'd also like to hear from others doing work with K-5 students in data literacy and associated skills, especially with professionally-collected real-world data.
Thanks for watching!
2
Discussion is closed. Upvoting is no longer available
I so appreciate your description of that sense of how emergent the reality was for trying to do research in classrooms ... that sense of "maybe in another month." And I love the solution of the Awesome Aquifer Kit! I'm also impressed with how you used video in order to still see what kids were doing and to conduct interviews. Wonderful creative thinking! I'm curious what you learned in the last year, or what methods you turned to, regarding the conduct of classroom-based research that you might retain going forward?
3
Discussion is closed. Upvoting is no longer available
Found helpful:
H Chad Lane
Remy Dou
Amy Busey
Jessica Sickler
Co-Presenter Principal
May 11, 2021 | 09:45 a.m.
Thanks, Leigh! I can speak from the perspective of the 1:1 clinical interviews -- where we adapted an in-person interview protocol to the virtual space, using collaborative platforms so that the kids and I could draw on the same visuals. I'm actually planning to adapt this technology and process for some pre/post assessment interviews when kids are back in schools. While I do really miss being side-by-side with the kids, it offers the opportunity to collect more data, from a much more geographically diverse sample of kids, with more flexibility, without massive amounts of travel (and all the carbon, time, expense, etc.) that entails. It's not the right fit for every research question, but I think it opened new possibilities and broadened my thinking. Necessity being the mother of invention, and all that.
I'm sure my colleagues will have thoughts about the curriculum development and testing lessons side of things!
1
Discussion is closed. Upvoting is no longer available
Found helpful:
Remy Dou
Ed Robeck
Lead Presenter Director of Education & Outreach
May 11, 2021 | 10:52 a.m.
Hi Leigh,
Thanks for your kind comments and your interest. Our findings are preliminary at this point, but the main takeaway so far is that (perhaps not surprisingly) the children are able to form highly reasonable conclusions using data representations (graphs, maps, tables). Their arguments often (again, not surprisingly) rely on their past experiences with the phenomena--or at least phenomena that they associate with the data. Their conclusions are not always correct from a scientific perspective, but they are reasonable in certain ways.
Regarding your second point--about what we might retain going forward--there are some strategies that I think we'll be trying when we can return to face-to-face teaching that are outgrowths of our forced virtual work. These have to do with how the data representations are formatted, for example, and the scaffolding that is used to help build understanding of those representations. The lack of physical models in the virtual settings meant that we were very deliberate about the representations. Would we have been as deliberate anyway? Perhaps. Using the representations in a more or less isolated way--without our preferred hands-on approaches--did let us hypothesize and test some approaches to their use that we might not have done otherwise. The power of the visualizations to evoke thinking is quite remarkable.
2
Discussion is closed. Upvoting is no longer available
Found helpful:
Remy Dou
Kimberly Elliott
Dionne Champion
Facilitator Research Assistant Professor
May 11, 2021 | 02:14 p.m.
Thanks for sharing this awesome project. I know that the focus of this video was more on the adjustments that your team had to make to your data collection process because of the pandemic, and there are a lot of lessons to be learned in that alone. However, I am interested in hearing more about how you decide which data sets to use as you develop your protocols. Are there certain types of data that seem to be more easily adaptable to the type of connections you want students to make?
2
Discussion is closed. Upvoting is no longer available
Found helpful:
Sabrina De Los Santos
Remy Dou
Jessica Sickler
Co-Presenter Principal
May 11, 2021 | 03:37 p.m.
Randy did a great job speaking from the curriculum side. I can add perspective from the first phase clinical interviews we did (to explore the natural assets children bring to thinking about data). In the very first study (still in-person), we presented them with commonplace data representations, stripping away the symbols of graphs and charts. They considered images representing intensity of rainfall, time/length of a rain event, and placement of a cup to collect rain to decide how different combinations of those variables would have led to the amount of water collected (more, less, or in-the-middle). They were complicated problems, with multiple, conflicting variables. And it was impressive how well kids were up to the challenge of wrestling with those variables, reasoning about an answer (even with no clear answer!), and they even raised about the data and how it was collected!
When we moved to interviews about more conventional representations, focusing on how well they could decode and read information about concepts -- a line graph (hydrograph), a bar graph (daily rainfall), and a rainfall map. Those were selected because they seemed most likely to be what the curriculum developers would use. One generalization I might make is that graphs containing a greater density of information (the hydrograph and the map) were more challenging for kids to draw meaning without scaffolding, compared to a bar graph. The instinct (habit?) of using graphs to read case values is strong, and those graphs make that impulse more difficult.
But scaffolding is, we believe, the way to get over that hurdle. When we started posing content questions (the barest bones of scaffolding) – “[From looking at this line graph of river height]: What do you think was going on with the weather on June 15?” It sometimes sparked “aha moments” to notice a part of the graph they didn’t before or hone their reasoning about what it was showing.
If I take the two parts of our Phase 1 research together, I think it emphasized that it’s not about which symbolic representation, but about a compelling problem, data (whether in a photo or a graph) to consider to solve it, and just-in-time scaffolding from an instructor to support them in getting unstuck.
2
Discussion is closed. Upvoting is no longer available
I love how you were able to adapt to the world being turned on its head! My project ended up being more an "accidental" solution to lockdown.
I'd love to hear more about what tools you used to pull off this transition.
1
Discussion is closed. Upvoting is no longer available
Found helpful:
Remy Dou
Jessica Sickler
Co-Presenter Principal
May 11, 2021 | 03:45 p.m.
This year was totally about creative problem-solving and accidental silver linings, right?? For adapting our 1:1 clinical interviews with students (the May-August part of our journey), we tried many solutions and ended up with a combination of platforms that met all of the constraints:
Google Meet for the videoconferencing (schools had suddenly banned Zoom as we were about to start)
Mural for the sharing of visuals, including: Interviewer control of what the student saw when, where they were focusing on the screen; Ability for both interviewer and child to see each other's cursors moving and interacting; Ability for both to draw, point, annotate the images to convey where they were looking, noticing, etc.
Screencast-o-matic for recording the entirity of the interaction (voice, video of faces, and collaboratively working with the digial materials).
When our curriculum team moved to testing actual lessons with small groups of students in-school and at home, it was a mix of Zoom/Google (whatever the school was using), screen-share from the presenter, an over-the-shoulder camera in the facilitator's office to capture the whole thing, and for some groups, the Awesome Aquifer kits that were mailed and distributed in advance by school partners. (Ed can weigh in with more detail on the logistics to all of this! I certainly missed something...)
2
Discussion is closed. Upvoting is no longer available
Found helpful:
Remy Dou
Kimberly Elliott
Randy Kochevar
Co-Presenter Director, Oceans of Data Institute
May 11, 2021 | 03:12 p.m.
Hi Dionne!
Thank you for your question. The process of developing curricula using authentic data is as much an art as it is a science. In this case, we began with a set analytical thinking outcomes that we hoped to support through the scaffolded learning experiences we were testing. These include:
Understanding the nature of data and their limitations:
Connecting data to real-world events or objects, where data representations are abstractions of those events or objects;
Considering data sources and limitations;
Shifting from viewing data as case values to data as an aggregate;
Asking and answering questions using data:
Mathematizing real-world problems and situations;
Making inferences and articulating levels of certainty;
Communicating with and about data:
Constructing narratives to explain patterns observed in data sets;
Engaging in argument with evidence from data.
We knew that we were interested in using data relating to rainfall and stream flow, in part because these are phenomena which seemed relatable to students at this age (grades 3-5). But the process of developing specific storylines and identifying datasets was iterative and organic. In the first phase of the project, we tested several different representations of these types of data including maps, histograms, and dot plots, to get a feel for the way students derived meaning from each of them.
For the second phase, which we are in now, we have experimented with a variety of different possible narratives, and evaluated each of them according to their potential to address the analytical thinking outcomes listed above. Once we settled on a topic that seemed to support our analytical thinking goals (periodic flooding of the Embarrass River Bridge in Sainte Marie, Illinois), we had to identify the specific datasets we would use, as well as the most effective ways to present them to the students. This has been an ongoing process, and as we test them with students, we continually modify and refine our approach based on the things we learn.
In general, we try to have the students work with datasets that are large enough and complex enough so there is not a single “correct” interpretation, or a definitive “right” or “wrong” answer. Instead, we try to support students in finding patterns or relationships within the data, making deductions from those observations, and using their findings as evidence to support their reasoning. This isn’t always easy, and we are hoping that this project will provide some insights about what kinds of facilitation practices work best for students at this age!
4
Discussion is closed. Upvoting is no longer available
Found helpful:
Sabrina De Los Santos
Remy Dou
Dionne Champion
Kimberly Elliott
Brian Dempsey
K-12 Teacher
May 11, 2021 | 03:42 p.m.
Data literacy is an essential skill for decision making in careers and everyday life. It is exciting to see an emphasis on this type of critical thinking at the elementary level. Thanks to everyone at Oceans of Data for this important work.
2
Discussion is closed. Upvoting is no longer available
Found helpful:
Remy Dou
Kimberly Elliott
Ed Robeck
Lead Presenter Director of Education & Outreach
May 11, 2021 | 06:53 p.m.
Hi Dionne,
In response to your question about datasets, Randy covered a lot of the curriculum design side well. I'll add that the project was really designed to explore ways that professionally collected can be used with young students. There is a lot of professionally collected data available online. We focused on geoscience data because both partnering organizations--EDC and AGI--have an interest there, which led us to the focus on hydrology. For example, the USGS website "USGS Current Water Data for the Nation" allowed us to navigate to hydrographs for rivers near our partner schools. We modified them a bit, primarily to take out perceptual noise, and shared them with students. We also downloaded precipitation data and constructed our own graphs from those data. Those are just a couple of examples. In making our adjustments, we tried to work within a framework of the way data are represented at the grade levels we are working, but with real-world data. It was interesting that we were even asked, "Are those real data?" as we interviewed children. It is no surprise that they are interested in working with data and information that relates to their own communities.
2
Discussion is closed. Upvoting is no longer available
Found helpful:
Remy Dou
Kimberly Elliott
Sabrina De Los Santos
Facilitator Research and Development Associate
May 12, 2021 | 01:29 a.m.
Fascinating work! Thank you for describing the ways you adapted once the pandemic hit and how you continued to engage learners remotely. I am curious to learn more about the role of the teachers in your program. In what ways did you collaborate with teachers and did you collect any data from them?
1
Discussion is closed. Upvoting is no longer available
Found helpful:
Remy Dou
Jessica Sickler
Co-Presenter Principal
May 13, 2021 | 10:08 a.m.
Adding to Ed's description of working with teachers -- Ed's prior working relationship with the school staff and teachers was extremely instrumental to us being able to push through the limitations of the pandemic. I can't stress the value of Ed's relationships and working with the schools enough! And we were all so appreciative of the teachers, the students, and the parents who took the leap with us and facilitated the work in the past year.
Beyond what Ed describes about interacting with teachers during lesson design this year, we didn't formally collect data from them (nor did we have explicit plans to). However, in our final phase -- when we implement the full 3-session classroom sequence, hopefully in person -- we do plan to conduct interviews with teachers about their observations of how the lesson went, what stood out to them in students' interactions and thinking, and anything they saw or heard after we left, etc.
1
Discussion is closed. Upvoting is no longer available
Found helpful:
Sabrina De Los Santos
Remy Dou
Facilitator Assistant Professor
May 13, 2021 | 12:07 a.m.
Comments: As mentioned in the comments above, this work is fascinating! I'm also very much uplifted by your presentations of research as messy in that it takes place in authentic contexts that aren't "sterile" sort to speak. And yet that hasn't hindered your ability to do something great that has positive impacts on students. Congrats!
Questions: While I'm eager to read more about your findings in the future, I resonate with Sabrina De Los Santos' comments and questions, and I'm also curious about the same.
1
Discussion is closed. Upvoting is no longer available
Found helpful:
Sabrina De Los Santos
Ed Robeck
Lead Presenter Director of Education & Outreach
May 13, 2021 | 08:37 a.m.
Hi Sabrina and Remy,
The question of teacher involvement is important to us. We recognize that we need to make use of both the formal knowledge base of teaching and the teachers' knowledge of their students. At first we intended to work directly with the teachers at the different phases of the research, working in partnership as we developed activities/protocols, and providing results of those phases both for their input and professional development. COVID changed that. We weren't able to meet with the teachers as intended and, especially at the beginning of the pandemic, there was not much appetite among the teachers for extra time online--the demands on them were already pretty intense. I'll skip over some of the "this led to that" scenarios and say that just as we found ways to interact with the students, we also found ways to interact with teachers and get their input about our protocols. This was facilitated by the fact that to move our work with the children to virtual modes, our protocols ended up being built on Google Slides that were screen-shared with the children through whatever platform their school system was using. Using Google Slides allowed us to share the slides with teachers before using them with children--Science, Mathematics, and English/Language Arts--and get their input. They offered great insights on a range of things, from types of data representations the students might be familiar with (or not) to key terms and pedagogical nuances (e.g., "notice" has a specific meaning among some of the teachers and students--as in the sense of open-ended observations--"What do you notice about the graph?") While as we finalize the protocols we don't always stay "in-bounds" in terms of curriculum and instructional standards, the teachers' input allows us to be intentional about the times we're pushing the children outside of their likely academic experiences in order to explore what thinking they're capable of. As summer approaches, discussions of results sharing and professional development are re-emerging.
1
Discussion is closed. Upvoting is no longer available
Great project and inspiringly frank sharing of challenges faced by circumstances and passion and effort to address.
Ed Robeck
Lead Presenter Director of Education & Outreach
May 14, 2021 | 08:28 a.m.
Thanks for the kind note, Andrew. It's been a ride. The team has been great. The real champions, though, have been the school folks. They've remained committed and supportive despite all their challenges. My respect for our K-12 school-based colleagues just keeps growing.
I love the student-centered approach your model takes to teaching data literacy. I work with the Passion-Driven Statistics model where we take the same approach. We believe that students from diverse educational, social and economic backgrounds deserve not only a seat in a classroom, but also a welcoming place at the table. Our goal is to increase the number and diversity of students exposed to meaningful and empowering data analysis experiences and to inspire the pursuit of advanced data-driven experiences and opportunities for everyone!
We have used this in higher education and with high school students in their course and as a summer boot camp. We have an upcoming workshop with middle school STEM teachers this summer that will implement next year in the classroom. We are excited to see our model being used by middle school students to complete project-based research using real data, asking original questions, and communicating methods and results using the language of statistics.
I see in the video the students are exposed to data, do they work with a software program to analyze their data?
Jessica Sickler
Co-Presenter Principal
May 14, 2021 | 01:17 p.m.
What a great program, Kristin! I'll have to check it out more. As Ed said, we don't have them working with software in this initial effort. I think partly because the intervention we're working toward in this project is much shorter duration than your examples (3 50-ish-minute classroom lessons). Our emphasis has been more on the analytical thinking skills -- how we can prompt them to occur, occur more deeply, and get students practicing the skills that Randy listed out in an earlier post.
FWIW, we identified this framework of analytical thinking skills because we found in earlier work with undergraduate students that these are things they struggled with when they encountered real, messy data for the first time in an oceanography class. For example, the skill of thinking critically about what a given data source doesn't tell you... it just wasn't something they'd been asked to do before. Coming up with an interpretation from data they could handle; but ask them to come up with an alternate interpretation... panic. Our hope is to start building the practice of these thinking skills (rather than the mechanics) at the outset, so that it's a natural habit of mind when the data gets even bigger and more complicated in high school and college.... or when they encounter a cool curriculum in middle school like yours!
With 9-12 graders we do a week book camp with maybe 12-15 hours of contact. We focus on descriptive statistics and graphing in SAS (free cloud based) with them developing a poster they present at the end. We have not done it with elementary but this summer I am running a workshop for middle school teachers that will implement next school year. I agree getting the younger students focused on analytical thinking is the way to go.
Ed Robeck
Lead Presenter Director of Education & Outreach
May 14, 2021 | 11:58 a.m.
Hi Kristen. I'm impressed with the description of your work and its motivations. I'll take a look at your video to learn more.
As for the software for use in analysis, that is not part of our design. It would be interesting to consider what grade 4 students might be able to do in that regard. Do you have experience at that level? I'll bring this up with the math teachers when I talk to them next.
The youngest we have worked with in a week boot camp is 9-12 graders but have plans for middle school aged next school year. We provide the code to copy and paste and they make minor changes with our guidance. We do a lot of one-on-one work in the boot camp.
Ed Robeck
Director of Education & Outreach
Welcome to our video about Streams of Data. This project will develop and test an approach that demonstrates how fourth grade students, supported by scaffolded learning experiences, can begin developing the analytical thinking skills necessary to make meaning from complex, large-scale, professionally collected data. The work will focus on analytical thinking as a subset of data literacy, specifically understanding the nature of data, asking and answering questions using data, and communicating with and about data.
We've been challenged to keep the project on track during the COVID-19 pandemic. Our video conveys some of our reflections on those challenges and how, with support from our partnering school systems, we've been able to keep the research going.
We'd like to hear about how others have navigated similar challenges with school-based research. We'd also like to hear from others doing work with K-5 students in data literacy and associated skills, especially with professionally-collected real-world data.
Thanks for watching!
Sabrina De Los Santos
Remy Dou
Leigh Peake
I so appreciate your description of that sense of how emergent the reality was for trying to do research in classrooms ... that sense of "maybe in another month." And I love the solution of the Awesome Aquifer Kit! I'm also impressed with how you used video in order to still see what kids were doing and to conduct interviews. Wonderful creative thinking! I'm curious what you learned in the last year, or what methods you turned to, regarding the conduct of classroom-based research that you might retain going forward?
H Chad Lane
Remy Dou
Amy Busey
Jessica Sickler
Principal
Thanks, Leigh! I can speak from the perspective of the 1:1 clinical interviews -- where we adapted an in-person interview protocol to the virtual space, using collaborative platforms so that the kids and I could draw on the same visuals. I'm actually planning to adapt this technology and process for some pre/post assessment interviews when kids are back in schools. While I do really miss being side-by-side with the kids, it offers the opportunity to collect more data, from a much more geographically diverse sample of kids, with more flexibility, without massive amounts of travel (and all the carbon, time, expense, etc.) that entails. It's not the right fit for every research question, but I think it opened new possibilities and broadened my thinking. Necessity being the mother of invention, and all that.
I'm sure my colleagues will have thoughts about the curriculum development and testing lessons side of things!
Remy Dou
Ed Robeck
Director of Education & Outreach
Hi Leigh,
Thanks for your kind comments and your interest. Our findings are preliminary at this point, but the main takeaway so far is that (perhaps not surprisingly) the children are able to form highly reasonable conclusions using data representations (graphs, maps, tables). Their arguments often (again, not surprisingly) rely on their past experiences with the phenomena--or at least phenomena that they associate with the data. Their conclusions are not always correct from a scientific perspective, but they are reasonable in certain ways.
Regarding your second point--about what we might retain going forward--there are some strategies that I think we'll be trying when we can return to face-to-face teaching that are outgrowths of our forced virtual work. These have to do with how the data representations are formatted, for example, and the scaffolding that is used to help build understanding of those representations. The lack of physical models in the virtual settings meant that we were very deliberate about the representations. Would we have been as deliberate anyway? Perhaps. Using the representations in a more or less isolated way--without our preferred hands-on approaches--did let us hypothesize and test some approaches to their use that we might not have done otherwise. The power of the visualizations to evoke thinking is quite remarkable.
Remy Dou
Kimberly Elliott
Dionne Champion
Research Assistant Professor
Thanks for sharing this awesome project. I know that the focus of this video was more on the adjustments that your team had to make to your data collection process because of the pandemic, and there are a lot of lessons to be learned in that alone. However, I am interested in hearing more about how you decide which data sets to use as you develop your protocols. Are there certain types of data that seem to be more easily adaptable to the type of connections you want students to make?
Sabrina De Los Santos
Remy Dou
Jessica Sickler
Principal
Randy did a great job speaking from the curriculum side. I can add perspective from the first phase clinical interviews we did (to explore the natural assets children bring to thinking about data). In the very first study (still in-person), we presented them with commonplace data representations, stripping away the symbols of graphs and charts. They considered images representing intensity of rainfall, time/length of a rain event, and placement of a cup to collect rain to decide how different combinations of those variables would have led to the amount of water collected (more, less, or in-the-middle). They were complicated problems, with multiple, conflicting variables. And it was impressive how well kids were up to the challenge of wrestling with those variables, reasoning about an answer (even with no clear answer!), and they even raised about the data and how it was collected!
When we moved to interviews about more conventional representations, focusing on how well they could decode and read information about concepts -- a line graph (hydrograph), a bar graph (daily rainfall), and a rainfall map. Those were selected because they seemed most likely to be what the curriculum developers would use. One generalization I might make is that graphs containing a greater density of information (the hydrograph and the map) were more challenging for kids to draw meaning without scaffolding, compared to a bar graph. The instinct (habit?) of using graphs to read case values is strong, and those graphs make that impulse more difficult.
But scaffolding is, we believe, the way to get over that hurdle. When we started posing content questions (the barest bones of scaffolding) – “[From looking at this line graph of river height]: What do you think was going on with the weather on June 15?” It sometimes sparked “aha moments” to notice a part of the graph they didn’t before or hone their reasoning about what it was showing.
If I take the two parts of our Phase 1 research together, I think it emphasized that it’s not about which symbolic representation, but about a compelling problem, data (whether in a photo or a graph) to consider to solve it, and just-in-time scaffolding from an instructor to support them in getting unstuck.
Remy Dou
Dionne Champion
Saira Mortier
I love how you were able to adapt to the world being turned on its head! My project ended up being more an "accidental" solution to lockdown.
I'd love to hear more about what tools you used to pull off this transition.
Remy Dou
Jessica Sickler
Principal
This year was totally about creative problem-solving and accidental silver linings, right?? For adapting our 1:1 clinical interviews with students (the May-August part of our journey), we tried many solutions and ended up with a combination of platforms that met all of the constraints:
I wrote a longer post about this process here, if you're super-interested.
When our curriculum team moved to testing actual lessons with small groups of students in-school and at home, it was a mix of Zoom/Google (whatever the school was using), screen-share from the presenter, an over-the-shoulder camera in the facilitator's office to capture the whole thing, and for some groups, the Awesome Aquifer kits that were mailed and distributed in advance by school partners. (Ed can weigh in with more detail on the logistics to all of this! I certainly missed something...)
Remy Dou
Kimberly Elliott
Randy Kochevar
Director, Oceans of Data Institute
Hi Dionne!
Thank you for your question. The process of developing curricula using authentic data is as much an art as it is a science. In this case, we began with a set analytical thinking outcomes that we hoped to support through the scaffolded learning experiences we were testing. These include:
Understanding the nature of data and their limitations:
Asking and answering questions using data:
Communicating with and about data:
We knew that we were interested in using data relating to rainfall and stream flow, in part because these are phenomena which seemed relatable to students at this age (grades 3-5). But the process of developing specific storylines and identifying datasets was iterative and organic. In the first phase of the project, we tested several different representations of these types of data including maps, histograms, and dot plots, to get a feel for the way students derived meaning from each of them.
For the second phase, which we are in now, we have experimented with a variety of different possible narratives, and evaluated each of them according to their potential to address the analytical thinking outcomes listed above. Once we settled on a topic that seemed to support our analytical thinking goals (periodic flooding of the Embarrass River Bridge in Sainte Marie, Illinois), we had to identify the specific datasets we would use, as well as the most effective ways to present them to the students. This has been an ongoing process, and as we test them with students, we continually modify and refine our approach based on the things we learn.
In general, we try to have the students work with datasets that are large enough and complex enough so there is not a single “correct” interpretation, or a definitive “right” or “wrong” answer. Instead, we try to support students in finding patterns or relationships within the data, making deductions from those observations, and using their findings as evidence to support their reasoning. This isn’t always easy, and we are hoping that this project will provide some insights about what kinds of facilitation practices work best for students at this age!
Sabrina De Los Santos
Remy Dou
Dionne Champion
Kimberly Elliott
Brian Dempsey
Data literacy is an essential skill for decision making in careers and everyday life. It is exciting to see an emphasis on this type of critical thinking at the elementary level. Thanks to everyone at Oceans of Data for this important work.
Remy Dou
Kimberly Elliott
Ed Robeck
Director of Education & Outreach
Hi Dionne,
In response to your question about datasets, Randy covered a lot of the curriculum design side well. I'll add that the project was really designed to explore ways that professionally collected can be used with young students. There is a lot of professionally collected data available online. We focused on geoscience data because both partnering organizations--EDC and AGI--have an interest there, which led us to the focus on hydrology. For example, the USGS website "USGS Current Water Data for the Nation" allowed us to navigate to hydrographs for rivers near our partner schools. We modified them a bit, primarily to take out perceptual noise, and shared them with students. We also downloaded precipitation data and constructed our own graphs from those data. Those are just a couple of examples. In making our adjustments, we tried to work within a framework of the way data are represented at the grade levels we are working, but with real-world data. It was interesting that we were even asked, "Are those real data?" as we interviewed children. It is no surprise that they are interested in working with data and information that relates to their own communities.
Remy Dou
Kimberly Elliott
Sabrina De Los Santos
Research and Development Associate
Fascinating work! Thank you for describing the ways you adapted once the pandemic hit and how you continued to engage learners remotely. I am curious to learn more about the role of the teachers in your program. In what ways did you collaborate with teachers and did you collect any data from them?
Remy Dou
Jessica Sickler
Principal
Adding to Ed's description of working with teachers -- Ed's prior working relationship with the school staff and teachers was extremely instrumental to us being able to push through the limitations of the pandemic. I can't stress the value of Ed's relationships and working with the schools enough! And we were all so appreciative of the teachers, the students, and the parents who took the leap with us and facilitated the work in the past year.
Beyond what Ed describes about interacting with teachers during lesson design this year, we didn't formally collect data from them (nor did we have explicit plans to). However, in our final phase -- when we implement the full 3-session classroom sequence, hopefully in person -- we do plan to conduct interviews with teachers about their observations of how the lesson went, what stood out to them in students' interactions and thinking, and anything they saw or heard after we left, etc.
Sabrina De Los Santos
Remy Dou
Assistant Professor
Comments: As mentioned in the comments above, this work is fascinating! I'm also very much uplifted by your presentations of research as messy in that it takes place in authentic contexts that aren't "sterile" sort to speak. And yet that hasn't hindered your ability to do something great that has positive impacts on students. Congrats!
Questions: While I'm eager to read more about your findings in the future, I resonate with Sabrina De Los Santos' comments and questions, and I'm also curious about the same.
Sabrina De Los Santos
Ed Robeck
Director of Education & Outreach
Hi Sabrina and Remy,
The question of teacher involvement is important to us. We recognize that we need to make use of both the formal knowledge base of teaching and the teachers' knowledge of their students. At first we intended to work directly with the teachers at the different phases of the research, working in partnership as we developed activities/protocols, and providing results of those phases both for their input and professional development. COVID changed that. We weren't able to meet with the teachers as intended and, especially at the beginning of the pandemic, there was not much appetite among the teachers for extra time online--the demands on them were already pretty intense. I'll skip over some of the "this led to that" scenarios and say that just as we found ways to interact with the students, we also found ways to interact with teachers and get their input about our protocols. This was facilitated by the fact that to move our work with the children to virtual modes, our protocols ended up being built on Google Slides that were screen-shared with the children through whatever platform their school system was using. Using Google Slides allowed us to share the slides with teachers before using them with children--Science, Mathematics, and English/Language Arts--and get their input. They offered great insights on a range of things, from types of data representations the students might be familiar with (or not) to key terms and pedagogical nuances (e.g., "notice" has a specific meaning among some of the teachers and students--as in the sense of open-ended observations--"What do you notice about the graph?") While as we finalize the protocols we don't always stay "in-bounds" in terms of curriculum and instructional standards, the teachers' input allows us to be intentional about the times we're pushing the children outside of their likely academic experiences in order to explore what thinking they're capable of. As summer approaches, discussions of results sharing and professional development are re-emerging.
Sabrina De Los Santos
Andrew Manches
Great project and inspiringly frank sharing of challenges faced by circumstances and passion and effort to address.
Ed Robeck
Director of Education & Outreach
Thanks for the kind note, Andrew. It's been a ride. The team has been great. The real champions, though, have been the school folks. They've remained committed and supportive despite all their challenges. My respect for our K-12 school-based colleagues just keeps growing.
Kristin Flaming
I love the student-centered approach your model takes to teaching data literacy. I work with the Passion-Driven Statistics model where we take the same approach. We believe that students from diverse educational, social and economic backgrounds deserve not only a seat in a classroom, but also a welcoming place at the table. Our goal is to increase the number and diversity of students exposed to meaningful and empowering data analysis experiences and to inspire the pursuit of advanced data-driven experiences and opportunities for everyone!
We have used this in higher education and with high school students in their course and as a summer boot camp. We have an upcoming workshop with middle school STEM teachers this summer that will implement next year in the classroom. We are excited to see our model being used by middle school students to complete project-based research using real data, asking original questions, and communicating methods and results using the language of statistics.
I see in the video the students are exposed to data, do they work with a software program to analyze their data?
Jessica Sickler
Principal
What a great program, Kristin! I'll have to check it out more. As Ed said, we don't have them working with software in this initial effort. I think partly because the intervention we're working toward in this project is much shorter duration than your examples (3 50-ish-minute classroom lessons). Our emphasis has been more on the analytical thinking skills -- how we can prompt them to occur, occur more deeply, and get students practicing the skills that Randy listed out in an earlier post.
FWIW, we identified this framework of analytical thinking skills because we found in earlier work with undergraduate students that these are things they struggled with when they encountered real, messy data for the first time in an oceanography class. For example, the skill of thinking critically about what a given data source doesn't tell you... it just wasn't something they'd been asked to do before. Coming up with an interpretation from data they could handle; but ask them to come up with an alternate interpretation... panic. Our hope is to start building the practice of these thinking skills (rather than the mechanics) at the outset, so that it's a natural habit of mind when the data gets even bigger and more complicated in high school and college.... or when they encounter a cool curriculum in middle school like yours!
Kristin Flaming
With 9-12 graders we do a week book camp with maybe 12-15 hours of contact. We focus on descriptive statistics and graphing in SAS (free cloud based) with them developing a poster they present at the end. We have not done it with elementary but this summer I am running a workshop for middle school teachers that will implement next school year. I agree getting the younger students focused on analytical thinking is the way to go.
Ed Robeck
Director of Education & Outreach
Hi Kristen. I'm impressed with the description of your work and its motivations. I'll take a look at your video to learn more.
As for the software for use in analysis, that is not part of our design. It would be interesting to consider what grade 4 students might be able to do in that regard. Do you have experience at that level? I'll bring this up with the math teachers when I talk to them next.
Best wishes,
Ed
Kristin Flaming
The youngest we have worked with in a week boot camp is 9-12 graders but have plans for middle school aged next school year. We provide the code to copy and paste and they make minor changes with our guidance. We do a lot of one-on-one work in the boot camp.
Further posting is closed as the event has ended.