Hi Josephine, thanks for your questions! Each of our school-contexts are slightly different, but we have been assessing student growth and engagement through pre and post surveys, interviews, and student artifacts. For example, looking at their ideas about what counts as data, how data is produced and used, competencies in interpreting graphs like scatterplots and dot plots/histograms. We also look closely at how students draw on the arts to make inferences based on data, going beyond just reading the data--making comparisons, hypotheses, explanations, thinking about their personal or local contexts and implications for people and communities. Many of these learning goals fit with the goals of critical data literacy. 

Thanks for your response, Anna. You're examining a rich set of student outcome measures and responses to your curriculum -- I look forward to learning from your findings!

We work with middle school students right now so it's hard to say how it might impact their future career paths, but we have found students excited to analyze data to investigate their communities, think about their own friendships, and highlight social issues they care about.

Oops! Posted my reply in the wrong spot. See my comments below.

The main way we are investigating this question is through enacting a codesign process to develop a curriculum (between researchers, math, art, ELA teachers). One trajectory we have observed is for educators to move from only planning activities in their own discipline, to trying out the tools of other disciplines, or even teaching ideas from other disciplines with support. Teachers may not feel confident teaching about data or art if it is outside of their expertise or may not see the integration as fundamental for learning in their discipline. Part of codesign is  building this confidence and discovery of how art and data literacy can reinforce goals in both disciplines through creative construction, iteration, and discussion. The scope and sequence of the curriculum will be different for each school depending on their goals and time. 

Thanks for the question! We were very interested in the student experience as well, so we conducted interviews with students after the projects wrapped up and also collected/analyzed artifacts from them as well.

And what we found was pretty broad, both in terms of developing data literacy knowledge/skills and socioemotional learning. So rather than try to summarize everything I'll point you to a few conference papers we have available. Lots more publications coming in the near future as well!

https://www.researchgate.net/publication/359196151_I_happen_to_be_one_of_478_Social-Emotional_and_Data_Reasoning_in_Middle_School_Students'_Comics_about_Friendship

https://www.researchgate.net/publication/360001008_What_Do_You_Meme_Students_Communicating_their_Experiences_Intuitions_and_Biases_Surrounding_Data_Through_Memes

https://www.researchgate.net/publication/359821240_Dancing_with_Data_Embodying_the_Numerical_and_Humanistic_Sides_of_Data

https://www.researchgate.net/publication/359196152_Postcards_and_photo_walks_Telling_community_data_stories_through_photography

Let me know if you have trouble accessing any of these.

Great question! So I'm not familiar with many studies that do a side by side comparison of arts-based and more traditional approaches to data science education that might speak to your question (if anyone else on the project wants to jump in with any, please do!), but what we are finding with our own research is that a lot of the humanistic elements of data science education described by Lee & Wilkerson come out through an arts-based approach approach: understanding how the data speaks to contexts, interpreting the data/applying it to real world situations, seeing oneself in the data, raising questions based on data interpretations, etc. If you want to read more about data art in data science education, I recommend tracking down the work of Rahul Bhargava, Catherine D'Ignazio, and Amy Stornaiuolo. All of them have done great work on this front.

As for what we found when it comes to dance-based data science curricula, you can find a recent conference paper from us on that topic here.

The schools we worked with formed a spectrum of how we were able to draw on arts education programs. The dance implementation was done in a context that had previous experience doing arts-integrated interdisciplinary projects (although data was new for them) and they had a dance teacher. Two others had a general arts courses that met once a week or so and one did not have a dedicated arts teacher so we drew on ELA for a kind of digital media arts focus. We selected the medium for the students through codesign with the teachers but one reason is that we wanted to support specific goals in art that may be different for illustration, visual storytelling, photography, dance, etc. The art mediums offered different ways of engaging with data. For example, in dance students embodied mathematical trends seen in graphical displays and offered space for interpretation through emotion. In comic-making students were given space to reflect on what statistics about friendship meant in their own lives and experiences, finding that personal connection to data. Photography helped students see connections between community data and what they see in their local neighborhoods and everyday routines. Photography offers space to communicate about social issues through both quantitative and qualitative data, to think about data in context, and visualize comparisons, patterns, associations. 

I would say there is a spectrum of how students were tasked with integrating art and data. One of the big impediments is just having more time to iterate on projects, respond to feedback or dig deeper into topics & questions they discover. In one school, access to computers became an issue and they also had to contend with unexpected interruptions due to testing. In photography, not all students were able to photograph their local neighborhoods the way they wanted to due to various impediments but could only take photos on the class-organized walks. In some contexts art is only once a week so being able to prioritize both art and math equally requires more time (e.g., 5 weeks vs. 2 weeks). It is also challenging to engage students in transforming large data sets if they don't have access to computers or even building new routines with data analysis tools is challenging to introduce for one unit. 

Thanks for the question! The main way we approach this aspect of the work is through coding the artifacts in different ways. Much like a teacher might use a rubric, we look for evidence of different components of data reasoning/literacy/science that students have included in the piece they produced. In coding the students comics, for instance, we looked for evidence of students engaging in different kinds of data reasoning: Descriptive Data Reasoning, Contextualizing the Data, Describing an Individual Experience in Relation to Data, Proportional Comparisons, Drawing Implications From the Data, and Listing further inquiries into the data (definitions and further analyses can be found in our paper here).

To measure learning, we've employed a few approaches along the ways with pre/post surveys and interviews being the most frequent tools we've used. This gives us a pretty holistic picture of what students learned in these units.

Thanks so much for these details and for the link to your paper - definitely helpful in thinking about what we can learn from students' expressive artifacts! 

Hi, thanks for your questions. Students reflected on their photos in terms of how they might tell a data story. For example, how do they help answer a statistical question and situate data in a local context (including setting the scene, showing a patter, contrast, or association, and representing the potential impact on people and places). They also reflected on composition strategies and how they used them to tell a story or express an emotion or idea. The focus on student inquiry prompted students to wrestle not only with challenges of getting out in their neighborhood to take photos, but coming up with investigative questions, thinking about how data they investigated might complement or support their photo essays, and rethinking inferences/perspectives based on what they saw in their local neighborhoods. The tools you mention sound really interesting, especially if we had a little more time for this project or students did an extension. I'll take a look at them. We would love to follow-up with you and think more about how we might collaborate.

Thanks for your question! And you're right, we definitely learned about co-designing interdisciplinary projects through this work!

Personally, I think the most important takeaway that we have landed on so far is that context matters in a big way. Student interests, school culture, administrative support, type or school, art forms, teacher backgrounds, data science education experience and focus, all of this plays into the success of the project and changes from school to school (and, in some cases, classroom to classroom). To this end, it was really important for us as researchers and co-designers to build in structures that had enough flexibility for teachers and students of all backgrounds to fit and grow within while also centering the goals we had as researchers and educators. I'm thinking here with Barab et al. (2004) when they talk about social commitments: decided up front what we, as co-designers want to hold to without dominating the conversation or overpowering the contextual nature of this work.

One other important lesson learned: be ready to support in a variety of ways. As we've been thinking through our process so far, we've spent a lot of time thinking about the roles of everyone involved. From the researcher side alone, we've at times been designers, advocates, data analysts, artists, students, teachers, and so many other things. And whenever we have taken on one of these roles, it's in support of the teachers implementing this work. Because the co-design process is so contextual, the supports teachers need have varied between schools and also changed as the project has grown. So we've been flexible and attentive, doing our best to help the teachers as they have implemented these projects.