Visualising nature through data embroidery: A hands-on workshop

Introducing data visualisation and storytelling without digital tools

JOANNE AMARISA

Founder, Data Garden Co.
Workshop Facilitato

Founder, Data Garden Co.
Workshop Facilitato

Founder, Data Garden Co.
Workshop Facilitato

MEI LEONG

Educator, Data Garden Co.
Project Lead and Facilitator

Educator, Data Garden Co.
Project Lead and Facilitator

Educator, Data Garden Co.
Project Lead and Facilitator

ABSTRACT

ABSTRACT

ABSTRACT

The field of data visualisation is increasingly dominated by digital tools that can rapidly transform datasets into charts, maps, and diagrams. While these tools offer speed and scalability, they can reduce users’ direct engagement with data, and create barriers to understanding how data is processed and communicated. We designed the “Data Stitching and Storytelling” workshop to demonstrate tactile and human-centered methods in the data-visual creation process. This hands-on, outdoor experience allows participants to more fully connect and engage with data. Each participant had an opportunity to observe and record data from their surroundings, analyse that data, and translate their findings into embroidered data representations. The workshop, geared toward young women but open to all ages and genders, combined computational thinking with hand-craft design. The hands-on process demystified data visualisation methods—particularly among participants who don’t regularly work with data. This white paper documents the workshop’s conceptual framing, materials, methods, and outcomes, and offers guidance for full or modular replication in public or academic settings.
The field of data visualisation is increasingly dominated by digital tools that can rapidly transform datasets into charts, maps, and diagrams. While these tools offer speed and scalability, they can reduce users’ direct engagement with data, and create barriers to understanding how data is processed and communicated. We designed the “Data Stitching and Storytelling” workshop to demonstrate tactile and human-centered methods in the data-visual creation process. This hands-on, outdoor experience allows participants to more fully connect and engage with data. Each participant had an opportunity to observe and record data from their surroundings, analyse that data, and translate their findings into embroidered data representations. The workshop, geared toward young women but open to all ages and genders, combined computational thinking with hand-craft design. The hands-on process demystified data visualisation methods—particularly among participants who don’t regularly work with data. This white paper documents the workshop’s conceptual framing, materials, methods, and outcomes, and offers guidance for full or modular replication in public or academic settings.
The field of data visualisation is increasingly dominated by digital tools that can rapidly transform datasets into charts, maps, and diagrams. While these tools offer speed and scalability, they can reduce users’ direct engagement with data, and create barriers to understanding how data is processed and communicated. We designed the “Data Stitching and Storytelling” workshop to demonstrate tactile and human-centered methods in the data-visual creation process. This hands-on, outdoor experience allows participants to more fully connect and engage with data. Each participant had an opportunity to observe and record data from their surroundings, analyse that data, and translate their findings into embroidered data representations. The workshop, geared toward young women but open to all ages and genders, combined computational thinking with hand-craft design. The hands-on process demystified data visualisation methods—particularly among participants who don’t regularly work with data. This white paper documents the workshop’s conceptual framing, materials, methods, and outcomes, and offers guidance for full or modular replication in public or academic settings.

INTRODUCTION

1.INTRODUCTION

1.INTRODUCTION

Data visual work today relies heavily on software and code to convert raw datasets into polished charts and diagrams. Platforms like Excel, Power BI, Tableau, Flourish, and Datawrapper, along with programming languages and libraries have become ubiquitous in workplaces thanks to their efficiency and scalability in processing large amounts of data. There is a significant barrier to entry for those who want to engage with data but do not know how to use these digital tools. Data work can feel exclusionary, particularly for historically underrepresented people in science and technology, including women who make up just 15% of STEM workers and 37% of STEM enrolments in Australia [1]. The problem is not just lack of technical knowledge; we have perceived, through our experience as data practitioners and educators, that when users “outsource” data processes to digital tools, they distance themselves from steps that would foster deeper understanding of the data itself. It is our concern that digital tools, while powerful, risk obscuring rather than supporting our interpretation of a dataset because they don’t account for the data’s flaws, biases, and surrounding context. We at Data Garden Co. (formerly The Data Garden Project), a nonprofit that provides data-based design, storytelling, and technology education for young women, developed the “Data Stitching and Storytelling” workshop in response to these challenges. The workshop took place on 15 February 2024 at MPavilion, located in the Queen Victoria Gardens, Melbourne, Australia, as part of the MPavilion 10 summer programme—a series of free public events that support MPavilion’s mission to champion architecture and design. In the course of 2.5 hours, we helped participants learn and engage with the fundamentals of data collection, analysis, and visualisation through an accessible and collaborative data crafting project. We began with a brief introduction to the field of data visualisation, and then guided participants as they collected data about nearby trees in the gardens. Following that exercise, participants used analytical thinking and their creative senses to encode their data in a visual format. Finally, they applied those encodings to a data crafting project that they could take home. Being in an outdoor environment and without any digital tools, participants relied on their natural data observation and data processing abilities throughout the data crafting process. Data crafting is a method of interacting with data and constructing visualisations with physical materials. It falls under the umbrella of data visualisation, which is the practice of encoding data variables into visual formats. And it is similar to data physicalisation, which encodes data variables into formats that are both visible and tangible, such as physical materials and 3D shapes. Data crafting is distinguished by the nature of the process and the final output: the process emphasises a playful approach to data in order to invite creativity and sensory exploration, and the output need not be perfectly and precisely constructed [2]. For purposes of this case study, we distinguish data crafting as a pedagogical method, and physical data visualisation as the end product. We chose embroidery as the data crafting medium for the workshop. Participants used threads and beads to encode the data they had collected about selected trees in the gardens [Figure 1]. The workshop was designed to host up to 15 participants and be inclusive to those with no prior experience in either data analysis or needlework. It offered a structured yet flexible approach: Participants could follow each step of the data pipeline, from data collection to stitching, while personalising choices at each step. This paper presents the “Data Stitching and Storytelling” workshop as a case study for inclusive data crafting education. We begin with the history of data crafting and its benefits to data practitioners today. Then we outline the development and execution of the workshop in detail, so that educators can replicate the workshop or create similar spin-off events with different crafting materials. Finally, we reflect on the participant outcomes and their subsequent feedback as well as our own takeaways from the experience.
Data visual work today relies heavily on software and code to convert raw datasets into polished charts and diagrams. Platforms like Excel, Power BI, Tableau, Flourish, and Datawrapper, along with programming languages and libraries have become ubiquitous in workplaces thanks to their efficiency and scalability in processing large amounts of data. There is a significant barrier to entry for those who want to engage with data but do not know how to use these digital tools. Data work can feel exclusionary, particularly for historically underrepresented people in science and technology, including women who make up just 15% of STEM workers and 37% of STEM enrolments in Australia [1]. The problem is not just lack of technical knowledge; we have perceived, through our experience as data practitioners and educators, that when users “outsource” data processes to digital tools, they distance themselves from steps that would foster deeper understanding of the data itself. It is our concern that digital tools, while powerful, risk obscuring rather than supporting our interpretation of a dataset because they don’t account for the data’s flaws, biases, and surrounding context. We at Data Garden Co. (formerly The Data Garden Project), a nonprofit that provides data-based design, storytelling, and technology education for young women, developed the “Data Stitching and Storytelling” workshop in response to these challenges. The workshop took place on 15 February 2024 at MPavilion, located in the Queen Victoria Gardens, Melbourne, Australia, as part of the MPavilion 10 summer programme—a series of free public events that support MPavilion’s mission to champion architecture and design. In the course of 2.5 hours, we helped participants learn and engage with the fundamentals of data collection, analysis, and visualisation through an accessible and collaborative data crafting project. We began with a brief introduction to the field of data visualisation, and then guided participants as they collected data about nearby trees in the gardens. Following that exercise, participants used analytical thinking and their creative senses to encode their data in a visual format. Finally, they applied those encodings to a data crafting project that they could take home. Being in an outdoor environment and without any digital tools, participants relied on their natural data observation and data processing abilities throughout the data crafting process. Data crafting is a method of interacting with data and constructing visualisations with physical materials. It falls under the umbrella of data visualisation, which is the practice of encoding data variables into visual formats. And it is similar to data physicalisation, which encodes data variables into formats that are both visible and tangible, such as physical materials and 3D shapes. Data crafting is distinguished by the nature of the process and the final output: the process emphasises a playful approach to data in order to invite creativity and sensory exploration, and the output need not be perfectly and precisely constructed [2]. For purposes of this case study, we distinguish data crafting as a pedagogical method, and physical data visualisation as the end product. We chose embroidery as the data crafting medium for the workshop. Participants used threads and beads to encode the data they had collected about selected trees in the gardens [Figure 1]. The workshop was designed to host up to 15 participants and be inclusive to those with no prior experience in either data analysis or needlework. It offered a structured yet flexible approach: Participants could follow each step of the data pipeline, from data collection to stitching, while personalising choices at each step. This paper presents the “Data Stitching and Storytelling” workshop as a case study for inclusive data crafting education. We begin with the history of data crafting and its benefits to data practitioners today. Then we outline the development and execution of the workshop in detail, so that educators can replicate the workshop or create similar spin-off events with different crafting materials. Finally, we reflect on the participant outcomes and their subsequent feedback as well as our own takeaways from the experience.
Data visual work today relies heavily on software and code to convert raw datasets into polished charts and diagrams. Platforms like Excel, Power BI, Tableau, Flourish, and Datawrapper, along with programming languages and libraries have become ubiquitous in workplaces thanks to their efficiency and scalability in processing large amounts of data. There is a significant barrier to entry for those who want to engage with data but do not know how to use these digital tools. Data work can feel exclusionary, particularly for historically underrepresented people in science and technology, including women who make up just 15% of STEM workers and 37% of STEM enrolments in Australia [1]. The problem is not just lack of technical knowledge; we have perceived, through our experience as data practitioners and educators, that when users “outsource” data processes to digital tools, they distance themselves from steps that would foster deeper understanding of the data itself. It is our concern that digital tools, while powerful, risk obscuring rather than supporting our interpretation of a dataset because they don’t account for the data’s flaws, biases, and surrounding context. We at Data Garden Co. (formerly The Data Garden Project), a nonprofit that provides data-based design, storytelling, and technology education for young women, developed the “Data Stitching and Storytelling” workshop in response to these challenges. The workshop took place on 15 February 2024 at MPavilion, located in the Queen Victoria Gardens, Melbourne, Australia, as part of the MPavilion 10 summer programme—a series of free public events that support MPavilion’s mission to champion architecture and design. In the course of 2.5 hours, we helped participants learn and engage with the fundamentals of data collection, analysis, and visualisation through an accessible and collaborative data crafting project. We began with a brief introduction to the field of data visualisation, and then guided participants as they collected data about nearby trees in the gardens. Following that exercise, participants used analytical thinking and their creative senses to encode their data in a visual format. Finally, they applied those encodings to a data crafting project that they could take home. Being in an outdoor environment and without any digital tools, participants relied on their natural data observation and data processing abilities throughout the data crafting process. Data crafting is a method of interacting with data and constructing visualisations with physical materials. It falls under the umbrella of data visualisation, which is the practice of encoding data variables into visual formats. And it is similar to data physicalisation, which encodes data variables into formats that are both visible and tangible, such as physical materials and 3D shapes. Data crafting is distinguished by the nature of the process and the final output: the process emphasises a playful approach to data in order to invite creativity and sensory exploration, and the output need not be perfectly and precisely constructed [2]. For purposes of this case study, we distinguish data crafting as a pedagogical method, and physical data visualisation as the end product. We chose embroidery as the data crafting medium for the workshop. Participants used threads and beads to encode the data they had collected about selected trees in the gardens [Figure 1]. The workshop was designed to host up to 15 participants and be inclusive to those with no prior experience in either data analysis or needlework. It offered a structured yet flexible approach: Participants could follow each step of the data pipeline, from data collection to stitching, while personalising choices at each step. This paper presents the “Data Stitching and Storytelling” workshop as a case study for inclusive data crafting education. We begin with the history of data crafting and its benefits to data practitioners today. Then we outline the development and execution of the workshop in detail, so that educators can replicate the workshop or create similar spin-off events with different crafting materials. Finally, we reflect on the participant outcomes and their subsequent feedback as well as our own takeaways from the experience.

Figure 1: Two completed embroideries from the workshop showcasing data observations about nine trees. The data is encoded as various stitch patterns, colours, and beads.

Figure 1: Two completed embroideries from the workshop showcasing data observations about nine trees. The data is encoded as various stitch patterns, colours, and beads.

Figure 1: Two completed embroideries from the workshop showcasing data observations about nine trees. The data is encoded as various stitch patterns, colours, and beads.

Background

Background

Background

2.1 Manual Data processes in the digital era

2.1 Manual Data processes in the digital era

2.1 Manual Data processes in the digital era

Long before spreadsheets and programming languages, humans used physical artefacts to collect, record, and process data. The Ishango bone, inscribed with tally marks over 20,000 years ago [3], and the Incan quipu, a knotted string system used to encode census and other inventory data [4], are two of many [5] examples of material methods used by early civilisations to collect and preserve data. These crafted representations of data allowed our ancestors to have sensory experiences (both visual and tactile) related to numerical data. The practice of transforming data into a physical object is known today as “data physicalisation”, a derivative of the broader term “data visualisation”. While physical objects are also visual, the distinction lies in how the data is converted or encoded. Jansen et al. [6] noted that “visualizations encode data in attributes of graphical marks such as points, lines, and areas. These attributes, for instance, size, color, or orientation are called visual variables. Similarly, physicalizations encode data in attributes of physical marks, and thus we call these attributes physical variables.” Humans today don’t rely on bones and strings to process data, but instead leverage digital and mechanical tools. Millions of people use software platforms like Excel, Power BI, Tableau, Flourish, or Datawrapper to generate digitised charts, diagrams, and maps. It is also becoming cheaper and faster [6] to create physicalisations with 3-D printers, laser cutters, and other tools. These automation tools have many benefits, as they can display data-based designs at speed and scale. Further, data processing capacities are becoming more sophisticated and efficient with the development of programming libraries and artificial intelligence. At the same time, automation tools risk displacing human sensory, emotional, and creative capacities that are exercised during hand crafting. Seghier cautions that the ways in which we handle big data—with increasing reliance on models and algorithms—is contributing to a “growing erosion of the human role in decision making and knowledge discovery process” [7]. Holmes is similarly critical, particularly of software tools that allow users to visualise data by selecting from a list of pre-set chart options. This ability, he states, has “removed human thinking” [8] from the data visual design process. Despite the overwhelming use of modern technology to process data and transform it into visual and physical formats, manual approaches that our early ancestors relied upon continue to have value in two key ways. First, data crafting gives us a richer understanding of the underlying data we are working with. Research in cognitive science supports the “embodied cognition” theory [9] whereby knowledge is constructed through the body’s movement, senses, and interaction with the environment. Like dancers who learn steps by marking (or practising with simplified movements) [10], the small act of notching tallies in a bone or tying knots in string or other motor experiences may help shape our cognition of the represented data. Therefore, there is a tradeoff between automation tools (which offer the user efficiency and scalability) and manual, hand-crafted approaches (which offer the user a more intimate cognitive engagement with the data). The data crafting process gives humans more agency to explore data at each step, from gathering the data to synthesising and visualising it. Second, data crafting embraces the inherent messiness of data. This is not always the case with automation tools, which produce precise outputs such as digitised charts and clean-cut physical objects. Audiences who come across such clean outputs may not recognise the ways in which the underlying data may be flawed or biased, nor how the output itself—whether a chart, diagram, or other construction—may be deceptive or misleading [11]. For those looking to work with data, there may be a belief that “data is austere—perfect—making it difficult to play with and explore” [2]. The intrinsic imperfections of manual crafting both reflect and normalise imperfections within the data and the human decisions that mold that data into a physical data visualisation. A growing number of data visual experts are advocating for closer, more critical examination of data as well as outputs that embrace its flaws. D’Ignazio and Klein highlight data projects that acknowledge data imperfections, uncertainties, and cultural entanglements in order to produce more ethical and inclusive data work [12]. Wu argues that “sensory and messy visualizations […] urge us to engage with them not as precise manifestations of a dataset, but instead prioritize an emotional connection and understanding” [13]. The principles of the “data humanism” movement spearheaded by Lupi [14] uphold a human-centered design approach that emphasizes the personal, contextual, and imperfect nature of data. Despite the prevalence of digital and automated data tools, there is still a place for the slower, more deliberate process of data crafting. As in centuries past, the manual process of working with data holds benefits to the practitioner, such as allowing a closer inspection and understanding of the data, and inviting more people to engage with data work by breaking data processes into discrete and manageable steps

2.2 Basis for workshop format and materials

2.2 Basis for workshop format and materials

2.2 Basis for workshop format and materials

Unlike formal instruction or self-guided experimentation, workshops provide a structured yet flexible space in which participants can encounter new methods, materials, and ideas with the guidance of facilitators and the support of peers. This structure is particularly beneficial for a group of participants who might be unfamiliar with the subject matter: Workshops can provide scaffolding while still offering room for personalisation, improvisation, and collaboration at each step of the process. This format also has precedent in the world of data crafting: One particular inspiration was the 2020 workshop held by Vladis et al. [2] that taught participants about data visualisation principles and multi-sensory modes of data encodings, and then facilitated hands-on data crafting projects. That workshop allowed participants to engage with data at their own comfort level. The emphasis was on process, not perfection, which encouraged participants to connect emotionally and materially with data literacy concepts that the facilitators had introduced. Our choice of embroidery as the core data-crafting medium was both conceptual and historical. The Jacquard loom, invented in 1804, is widely regarded as a precursor to modern computing. Its use of punch cards [15] to control yarn patterns foreshadowed early programming logic that then became the building blocks for our modern-day digital data systems. At the same time, embroidery is a type of labour traditionally classed as “women’s work.” In 19th century Australia, embroidery was maintained as a social and domestic tradition from England and “was considered an essential skill” for wives and mothers as well as those seeking employment opportunities [16]. Girls as young as five years old and from all social classes sewed samplers to demonstrate their abilities and exemplify their agency [17]. This history resonated with our mission at Data Garden Co. to build agency in communities, especially those of women, for artistic expression and for learning data, code, and computational thinking through visual and creative tools [18]. Embroidery facilitates our mission of bringing people closer to data and helping them engage with it on a deeper level. For one thing, it has been suggested that embroidery is well suited for recording personalised data [19], which can be more engaging than data obtained from a third party. (The workshop focused on data that reflected each participants’ personal observations of trees.) Also, embroidery requires that participants take a measured and deliberate approach to data processing. As textile artist Jordan Cunliffe notes, the field of data visualisation is “fast-paced and ever-changing, yet when paired with such a traditional skill as hand embroidery, it becomes laborious and slow and meditative.” [20] We drew inspiration from contemporary textile artists such as Cunliffe and Ahree Lee because their work encodes quantifiable — and often countable — data using materials such as threads and beads. For Cunliffe, the act of measuring, counting, and stitching is a means of “meticulously and accurately” recording data, such that 10,000 data points will result in 10,000 stitches [20]. Finally, manual embroidery also has intrinsic limitations and challenges, such as rendering straight lines or applying the appropriate tension so stitches are not too tight or too loose. This imprecision naturally embraces the “messiness” of data that stems from factors such as estimated measurements, or subjective observations or interpretations.

OBJECTIVES

1.INTRODUCTION

1.INTRODUCTION

The topline purpose of this workshop was to spread the principles of the data humanism movement [14], which fosters deeper understanding and more meaningful personal experiences with data. Our more specific objectives for the participants mimic the 2020 workshop held by Vladis et al. [2], which, in turn, references the six levels of cognitive learning in Bloom’s Taxonomy of Educational Objectives [21]: - Knowledge: Recall that data exists all around us and is not inherently digital. Visualising information can take many different formats. - Comprehension: Understand the stages of creating a data visualisation (data collection, analysis, synthesis, sketching, and representation) and how these stages can be conducted through sensory experiences and human decision making. - Application: Gather observations from the surrounding environment and translate them into hand-crafted physical data visualisations. - Analysis: Categorise data into sets of qualitative and/or quantitative variables and match them to visual variables such as colour, size, shape, etc. Synthesis: Design and create a hand-crafted physical data visualisation using provided physical (embroidery) materials. - Evaluation: Reflect on the creative process of data representation, recognising how human perception and data context shapes how we interpret and communicate data.

Planning and logistics

1.INTRODUCTION

1.INTRODUCTION

Planning for the “Data Stitching and Storytelling” workshop took place over a period of three months. Our organisation had already begun conceptualising a stitching-based data visualisation workshop when the opportunity to participate in MPavilion’s MMeets 2024 summer programming emerged. MPavilion had previously held workshops on fibre crafts and textile arts, so we believed our workshop would be a good fit for their mission. Upon acceptance into the MMeets series, we adapted the structure and content of the workshop to align with MPavilion’s guidelines and site-specific requirements, such as keeping the event to 2.5 hours and refraining from using digital technology. These parameters shaped the final format of the day. The workshop was held within Queen Victoria Gardens, with the MPavilion itself serving as the central site. Designed by renowned Japanese architect Tadao Ando, the pavilion is a concrete structure that offers a tranquil, shaded, and acoustically insulated environment. Given the seasonal timing of the event in February, weather conditions were favourable and no rain contingency plan was required. The event was promoted primarily through social media channels, though interest was also driven by MPavilion’s established reputation and the popularity of its summertime programming.

4.1 Choosing data to reflect the natural setting

4.1 Choosing data to reflect the natural setting

4.1 Choosing data to reflect the natural setting

The park setting offered an array of natural elements (plants, rocks, trees, and water features, for instance) that could serve as potential subjects for data collection. After visiting the site, we decided to focus the workshop on trees surrounding the pavilion. The City of Melbourne’s Open Data Portal contains a dataset on trees [23] around the city, with information such as tree name, scientific name, genus, date planted, diameter, and geographic location. From this dataset, we were able to extract information about the trees surrounding the pavilion. Although workshop participants could use this data, we also wanted them to collect their own observational data about the trees and juxtapose the two; we drew inspiration from Hradil’s workshop [24] at the University of Pennsylvania that demonstrated how short, reflective nature walks and subjective observations about trees can yield personally meaningful datasets. We initially selected 18 trees, all a short walking distance from the pavilion and near to pedestrian paths. However, after several preparatory run-throughs, we narrowed the list to nine trees (by eliminating some duplicate species) to ensure the participants weren’t rushed to collect data for each tree within the timeframe or overwhelmed by the amount of data they’d have to translate to the embroidery.

4.2 Developing worksheets for beginners

4.2 Developing worksheets for beginners

4.2 Developing worksheets for beginners

We created five worksheets to help participants record observational data about each tree, translate that data into visual variables, and sketch their visualisation before stitching. The first four of these worksheets were included in a booklet [Appendix A, Figures A3-A5] and the final one, for sketching, was distributed separately [Appendix B, Figures B1-B3] after participants had chosen an embroidery template. The booklet also contained a reference sheet [Figure A3] to provide additional guidance. This sheet included a tree map (trees were identified as No. 1 through No. 9) in relation to the pavilion. It also included a table with the two variables from the City of Melbourne’s Open Data dataset [23] (tree common name and diameter at breast height) to help participants identify each tree as they walked. It was up to participants whether or not to incorporate this information into their visualisations. Each worksheet focused on one step of the data process, effectively breaking the workshop into manageable stages: Ideation worksheet [Figure A3]: This worksheet prompted participants to brainstorm questions for data collection (‘Does this tree provide shade?’ or ‘How climbable is this tree?’ etc.), and measurements or categories to answer those questions (‘yes/no’ or ‘climbable, maybe climbable, not climbable,’ etc.). Finally, it prompted them to think about ways they might visually represent those categories using embroidery materials (‘a line of stitches,’ or ‘a bead,’ etc). Field data collection sheet [Figure A4]: Organised as a table, this worksheet listed the names of the nine trees in the first column and provided three additional columns for participants to record their own measurements or observations for up to three variables. Encoding worksheet [Figure A5]: This exercise, which we adopted from Posavec’s manual encoding methods [25], helped participants plan their data encodings. It invited them to list their tree variables on the left side of the page and possible visual attributes (thread colour, bead colour, stitch pattern, stitch length, etc.) on the right side. They could then draw a line to connect each variable on the left to a visual attribute on the right. Stitching legend worksheet [Figure A5]: This worksheet aided participants in developing a visual key. They could either sketch how they would stitch each variable or tape a sample of the material (thread or bead) they would use to encode each variable. This reinforced the idea that every stitched element would represent a data point. Sketching sheet [Figures B1-B3]: Participants could sketch their embroideries on a loose paper template before stitching. We made sure to have one completed booklet on hand for reference. This enabled participants to see an example of each finished worksheet as they completed their individual worksheets

4.3 Prototyping crafting materials

4.3 Prototyping crafting materials

4.3 Prototyping crafting materials

We tested materials to identify effective and user-friendly embroidery tools, keeping in mind that participants may not have prior needleworking experience. The prototyping process involved testing various embroidery fabrics, needles, threads, and beads. We evaluated factors such as whether the thread could be easily inserted through the eye of the needle without splitting, and whether the beads could be strung on to the thread. Another consideration was the number of colour options; we wanted enough variety in thread and bead colours for participants to be creative but not be overwhelmed by the choices. The thread and bead colours also had to have corresponding marker colours for the sketching phase. Finally, we designed three visualisation templates (a map, a donut shape, and a dot grid) [Figure A4] from which participants could choose one. Because freehand embroidery can be challenging, we wanted the participants to see their chosen template directly on the fabric. To accomplish this, we printed the templates on Solvy, a water-soluble adhesive sheet that participants could affix to the fabric on their embroidery hoop [Figure 2]. After the workshop (at home), they could place their embroidery in cold water to dissolve any part of the underlying template that was still visible.

MATERIALS

1.INTRODUCTION

1.INTRODUCTION

Displays and exemplars [Figure 3]: - 2 Completed exemplar data embroideries. - 3 A3-size posters showing completed data sketches of the map, donut, and dot-grid templates, along with a key matching the data to visual variables. - 3 Tabletop A3 easels to display the posters. - 2 Aida cloths showing various stitching and beading techniques, for creative inspiration. - 1 completed booklet for reference. Record, Map and Capture in Textile Art: Data Visualization in Cloth and Stitch by Jordan Cunliffe. (Book on display and available to browse for data embroidery examples and inspiration.) Worksheet materials and templates: - 15 Booklets [Appendix A] presented as an A4-size folded booklet containing worksheets and other reference materials. Booklets were printed on A3 duplex (2-up saddle stitch imposition). - 30 True-to-scale (110mm diameter circle) templates for sketching the embroidery design (10 per design template), printed on A4-size paper, and with annotations to guide the participants on visualising each tree. - 30 True-to-scale (110mm diameter circle) templates printed on Solvy adhesive-backed substrate. (10 for each template, and already pre-cut.) Multipack black fibretip pens. - 18 Packs of coloured marker pens, - 12 colours matching thread and bead colours. - 2 Rolls of tape. Embroidery Materials - 15 Embroidery hoops with wooden screw, 130mm diameter. - 15 Units of Aida cloth 14-count cross stitch fabric, pre-cut to 180mm x 180mm. - 15 Tapestry needles, size 22. Olympus Sashiko threads, 20m cotton skeins in 12 colours. Glass seed beads in segmented containers, pre-sorted by colour. (Bead eye large enough to fit over size 22 tapestry needle.) - 2 A4-size felt sheets cut into 15 units to store needles. - 1 Unpicker. - 3 Needle threaders. - 9 Craft scissors. NOTE: We pre-cut the Sashiko threads into approximately 1-1.5m lengths and wound them onto individual plastic bobbins using a thread-winder tool. These individual units made it easier for participants to “grab-and-go” when they were ready to start stitching and to take home extra thread if they hadn’t completed their stitching by the workshop’s end.

Figure 3: A table displaying poster-size exemplars of the completed encoding worksheet and the embroidery template sheet, as well completed sample embroideries for each template. 

WORKSHOP EXECUTION

1.INTRODUCTION

1.INTRODUCTION

Day-of setup Chairs in MPavilion were arranged such that participants could sit together around a large table but have enough space to work individually. At each seat we placed a booklet, a needle in a felt square, one embroidery hoop pre-loaded with fabric, and a pen. Markers, scissors, threads, beads, and other communal supplies were placed strategically around the table. Welcome and icebreaker 8 minutes Once participants had been registered and seated at the table, we spoke briefly about our organisation and shared the names and roles of the facilitators and volunteers. Then we asked the participants to introduce themselves to the participants seated around them. We also ensured that each participant had a booklet and a pen. Workshop introduction 7 minutes We showed the pre-fabricated examples so participants could see what they would be making. We used these examples to introduce the concept of data crafting, explaining that the embroideries were crafted from data about trees in the surrounding environment. From there, we introduced the five steps of crafting a data visualisation: 1) Ideate: Come up with an idea, which often starts by asking questions that can be answered through data. 2) Collect the data: Find the data or build a dataset to answer those questions. 3) Parse the data: Categorise and/or clean up the data to make it easier to work with. 4) Visually map the data: Define how the input data will be represented (or encoded) as a visual output. 5) Create: Transfer those encodings into a physical visualisation. It was important for participants to understand that these steps were executable without any digital technology. As such, we briefly touched on what the data humanism movement is, and that, as people, we can use our sensory experiences to recognise data,collect it, and tangibly work with it in our everyday lives. Finally, we gave an overview of the worksheets in the booklets [Appendix A], noting that each worksheet would help facilitate the five steps. Ideation worksheet 5 minutes Participants filled in their ideation worksheet [Figure A3], which prompted them to think about tree characteristics that they could observe on their walk (‘Does this tree provide shade?’ or ‘How climbable is this tree?’ etc.) and how they might measure or categorise those characteristics (‘yes/no’ or ‘climbable, maybe climbable, not climbable,’ etc.). It also prompted them to think about ways they might visually represent those categories using embroidery materials (‘a line of stitches,’ or ‘a bead,’ etc). Tree walk and field data worksheet 20 minutes As a group, the participants walked around the gardens, stopping at each of the nine trees labeled on the map in their booklets [Figure A2]. On this map, trees were identified as No. 1 through No. 9. The booklets also included a data table [Figure A3] with two variables from the City of Melbourne’s Open Data dataset [23] (tree common name and diameter at breast height) to help participants identify each tree as they walked. At each tree, participants recorded information corresponding to the variables they had chosen to track. They used their field data worksheet [Figure A4], which offered a structured data table for them to fill in. We encouraged participants to observe two or three variables that would be consistent across the trees. Debrief and context 8 minutes Upon returning to the pavilion, we recapped where we were in the five-step data process: Participants had completed the ideation and data collecting in steps 1 and 2 and now needed to start working with that data in steps 3 through 5. We explained the concept of mapping (or encoding) data to visual variables to create a legend. For clarification and inspiration, we showed Jordan Cunliffe’s [20] data embroidery project that conveyed data about flowers spotted on a walk: Cunliffe recorded the colours of the flowers and then used those colours — and the order in which they were recorded — as a colour pattern for the threads. This example illustrated algorithmic thinking: each of Cunliffe’s flowers was translated into a stitch. Then, we returned to our exemplar embroideries to reiterate how the tree data could be encoded into stitches. We showed that different choices, such as stitch type or direction, stitch length, stitch colour, and bead colour, were all possible types of encodings. Picking a template 2 minutes We provided the participants with three templated design options (a map, a donut shape, and a dot grid) to craft their embroidery [Figure A4]. We asked each participant to select one of the designs for their project though we also stated that they could make their own design. We explained how the areas (numbered 1 to 9) on the templated designs corresponded to the trees they had seen on the walk. Parsing data using the encoding worksheet and the stitching legend worksheet 10 minutes After briefly reviewing the materials that were available, we asked participants to fill in their encoding worksheet [Figure A5], writing their data categories on the left side of the page and possible visual attributes (thread colour, bead colour, stitch pattern, stitch length, etc.) on the right side. Participants effectively cleaned their data during this exercise, by grouping their written observations into distinct categories. They could then draw a line to connect each variable on the left to a visual attribute on the right. It was up to participants whether or not to incorporate information from the City of Melbourne’s Open Data dataset (tree common name and diameter at breast height) into their visualisations. Once they had “mapped” their data variables to their visual encodings, they were invited to create legends using the legend worksheet [Figure A5]. They could sketch their legends with markers and/or tape samples of materials (such as threads and beads) that would make up their legends to the paper. Sketching worksheet 10 minutes Using both the field data worksheet and the legend worksheet as reference, the participants used coloured markers to sketch their embroideries onto a paper template [Figure 4]. Stitching 75 minutes In tandem with our demonstration, participants adhered their chosen template onto the topside surface of the cloth loaded in the embroidery hoop using a Solvy adhesive sheet. Finally, they threaded their needles and started stitching. Facilitators were available to help with technical challenges such as threading, knotting, or beading. Share and conclusion 5 minutes In the final minutes, participants could briefly share their findings or outcomes with others at the table. We closed with a final reiteration of the workshop’s purpose: to demonstrate that embroidery is one medium in which we can express data, and that we can always be more mindful of the data that exists in nature around us.

Figure 4:Sketches of the embroideries, with encoding and legend notes, created with coloured markers. 

WORKSHOP OUTCOMES

1.INTRODUCTION

1.INTRODUCTION

The “Data Stitching and Storytelling” workshop was a free event open to the public, but with a maximum of 15 participants. Ahead of the workshop, 15 people had registered and, of them, seven attended. An eighth participant who had not registered also joined. Among those who attended, experience levels varied widely. Two participants reported prior experience with embroidery, while one had never threaded a needle before. Most participants had some ad-hoc sewing familiarity, such as replacing a button. On the data side, three attendees came from science-related fields and were familiar with working with data in digital formats, though none had previously engaged with data through creative or material processes. This is an overview of participants’ workshop experiences and their final embroidery projects based on our observations and based on feedback we received after the event through a short questionnaire that three attendees completed. Overall, participants embraced the principles of data humanism with surprising ease. Several participants demonstrated an intuitive ability to record observational data using relational comparisons (e.g. “This bark is rougher/smoother”) and embodied measurement strategies (e.g. “The branches begin at twice my height”). Participants worked individually, coming up with their own tree variables, though they shared their variables and observation notes with fellow participants along the walk. One participant recorded data about the walk itself, recording, for instance, the number of participants that touched each tree. Once back in the pavilion, participants completed the worksheets and stitching at varying speeds. For instance, one participant who recorded a particularly large number of variables on the walk took nearly 20 minutes (double the expected time) to parse the data for encoding. When it came to stitching, the map-based template [Figure A4] was the most popular of the three options. Participants gravitated toward natural tones in their palettes—greens, browns, and greys—often matching their thread colours to elements of the surrounding environment, such as the concrete pavilion or leafy canopies. In terms of timing and logistics, we encountered two bottleneck points during the workshop: First, during the data parsing stage, several participants needed help turning their field notes into usable data for encoding. We found that these participants either hadn’t quantified their observations or, in the case of qualitative data, had too many disparate categories that needed to be consolidated. Second, at the start of the stitching phase, multiple participants often required assistance with embroidery basics like threading the needle or tying the first knot. At both points, multiple participants needed assistance simultaneously and could not proceed until they had been assisted. The final embroidery pieces reflected genuine engagement and thoughtful translation of data into a physical artefact. Some participants did not finish stitching by the end of the workshop, but were provided materials to complete their work at home. The follow-up questionnaire responses were largely positive; all three of the participants who responded strongly agreed that the workshop was engaging and that the worksheets and visual guides were helpful. On a more critical note, they would have liked the event to be longer. From their responses, we believe that they found the workshop to be educational. Participants noted that they “loved being able to collect the data myself” and appreciated the freedom to “pick data that was relevant to me/my interests.” One respondent shared, “I learned about data visualisation and the creative process of visualising data,” while another cited “learning a new area of data collection, interpretation and analysis.”

SUGGESTIONS FOR FUTURE WORKSHOPS

1.INTRODUCTION

1.INTRODUCTION

Based on the outcomes of the workshop, we met the educational objectives that we had established at the outset: to support participants as they explore data in a personal, “more human” way. Even our participants who had no prior experience in embroidery or data analysis were able to engage meaningfully with the process. This proved to us, as an inclusive educational organisation, that data crafting provides a low-barrier point of entry into data exploration. However, should we (or others) replicate the workshop, we would consider a few modifications. Our ratio of seven participants to five facilitators was generous, but we felt that hands-on support was stretched at key moments. We would suggest more facilitators for a larger group and also more time for the workshop—perhaps four hours minimum. This would enable facilitators to check in with participants between each step, and grant participants more opportunity to present their work to others at the end. In particular, we would suggest a longer check-in period before the data walk to ensure that participants have clearly defined just two or three measurable variables and that they have ideas about how to measure or categorise those variables. We believe this would make the encoding steps more seamless. Another recommendation concerns the design stages. We believe the multi-worksheet system was helpful to break down the process; however, some participants expressed confusion about moving from one sheet to the next, particularly when mapping data variables to visual elements, and then using those visual elements in their legends and sketches. To address this issue, facilitators should make it clear that the data steps (and their corresponding worksheets) are not always linear, and that participants should feel free to look back at their completed worksheets from an earlier step at any point during the process. In addition, the steps might be simplified in the future by eliminating the legend worksheet, which would help participants get to sketching and crafting quicker. Alternatively, the step-by-step process could be reinforced through poster-size, pre-completed worksheets, which participants could reference periodically in order to better understand the purpose of each worksheet and how to complete it.

CONCLUSION

1.INTRODUCTION

1.INTRODUCTION

The “Data Stitching and Storytelling” workshop invited participants of all levels to explore the fundamentals of data collection, analysis, and visualisation through a hands-on embroidery project. Participants came up with questions about trees in Melbourne’s Queen Victoria Gardens, recorded characteristics of those trees (data variables) to answer those questions, then visually encoded that information into a legend. Finally, they used threads and beads to translate those encodings onto fabric. In a digital era where data work is strongly associated with coding, spreadsheets, and charting software, this workshop demonstrated that tactile, craft-based projects can foster computational thinking that feels natural. For educators who are exploring new ways to promote data exploration, this case study demonstrates how data crafting can support inclusive and humanistic approaches to data learning.

ACKNOWLEDGEMENTS

1.INTRODUCTION

1.INTRODUCTION

Funding for the workshop was provided by Data Garden Co. and MPavilion. Many people made this workshop possible. We’d like to thank Bayu Sasono, Sidney Taylor, and Catherine Lewis, who volunteered as additional facilitators on the day, and Larissa Serafina of Data Garden Co. who assisted in creating promotional content for the event. Special thanks to our mentor Karen Ann Donnachie who helped us develop ideas for our data encoding systems and templates, as well as a personalised approach to the data collection exercise while leveraging the City of Melbourne’s Open Data resource. We appreciate the opportunity to work with MPavilion. We would like to thank the organisation for funding support, offering a serene space to conduct the workshop, and their commitment to facilitating community-centred experiences that promote the arts and nature.

References

1.INTRODUCTION

1.INTRODUCTION

[1] Australia Department of Industry, Science and Resources. 2024. Department of Industry, Science and Resources: The state of STEM gender equity in 2024. Retrieved from https://www.industry.gov.au/news/state-stem-gender-equity-2024 . [2] Nathalie Alexandra Vladis, Aspen K. Hopkins, and Arvind Satyanarayan. 2020. Data crafting: Exploring data through craft and play. Retrieved from https://vis.csail.mit.edu/pubs/data-crafting.pdf . [3] Smithsonian Institution. 2024. Smithsonian National Museum of Natural History: What does it mean to be human? Retrieved from https://humanorigins.si.edu/evidence/behavior/recording-information/ishango-bone . [4] Museo Larco. 2018. Museo Larco: Inca Quipus. Retrieved from https://www.museolarco.org/en/exhibition/permanent-exhibition/online-exhibition/textiles-from-ancient-peru/inca-quipus/ [5] Pierre Dragicevic and Yvonne Jansen. 2012. List of physical visualizations. Retrieved from https://dataphys.org/list/ . [6] Jason Alexander, Yvonne Jansen, Kasper Hornbæk, Johan Kildal, and Abhijit Karnik. 2015. Exploring the challenges of making data physical. In CHI EA '15: Proceedings of the 33rd Annual ACM Conference Extended Abstracts on Human Factors in Computing Systems, April 18 - 23, 2015, Seoul, Republic of Korea. ACM Inc., New York, NY, 2417 - 2420. https://doi.org/10.1145/2702613.2702659 [7] Mohamed Seghier. 2021. An active human role is essential in big data-led decisions and data-intensive science. F1000Research 10 (November 2021), 1127 pages. https://doi.org/10.12688/f1000research.73876.1 [8] Nigel Holmes. 2022. Joyful Infographics: A Friendly, Human Approach to Data. A K Peters/CRC Press, Boca Raton, Florida, USA. https://doi.org/10.1201/9781003222361 [9] Anna M. Borghi and Felice Cimatti. 2010. Embodied cognition and beyond: Acting and sensing the body. Neuropsychologia, 48, 3 (February 2010), 763-773. https://doi.org/10.1016/j.neuropsychologia.2009.10.029 [10] Edward Warburton, Margaret Wilson, Molly Lynch, and Shannon Cuykendall. 2013. The cognitive benefits of movement reduction evidence from dance marking. Psychological Science, 24, 9 (July 2013), 1732-1739. https://doi.org/10.1177/0956797613478824 [11] Aspen Hopkins, Michael Correll, and Arvind Satyanarayan. 2020. VisuaLint: Sketchy In Situ Annotations of Chart Construction Errors. Computer Graphics Forum, 39, 3 (June 2020), 219-228. https://doi.org/10.1111/cgf.13975 [12] Catherine D’Ignazio and Lauren F. Klein. 2020. Data Feminism. The MIT Press, Cambridge, Massachusetts, USA. https://doi.org/10.7551/mitpress/11805.001.0001 [13] Shirley Wu. 2024. Senses and sentiment: When data is too emotional for the screen. Nightingale: The Journal of the Data Visualization Society, 4 (March 2024), 66-71. https://nightingaledvs.com/senses-and-sentiment/ [14] Giorgia Lupi. 2017. Data humanism, the revolution will be visualized. Retrieved September 15, 2025 from https://giorgialupi.com/data-humanism-my-manifesto-for-a-new-data-wold . [15] The Editors of Encyclopaedia Britannica. 2022. “Jacquard loom". Encyclopedia Britannica. Retrieved from https://www.britannica.com/technology/Jacquard-loom [16] State Library of South Australia. A lifetime of stitching: Historic needlework representing the lives and work of South Australian girls. Retrieved September 15, 2025 from https://stories.slsa.sa.gov.au/a-lifetime-of-stitching/ [17] Catherine Gay. 2019. All life and usefulness: Girls and needlework in 19th-century Victoria. University of Melbourne Collections, 25 (December 2019), 34-40. https://museumsandcollections.unimelb.edu.au/__data/assets/pdf_file/0005/3272333/10-Gay-Needlework-25.pdf [18] Joanne Amarisa and Arran Ridley. 2024. Introducing girls to code, one flower at a time. Nightingale: The Journal of the Data Visualization Society. Retrieved September 15, 2025 from https://nightingaledvs.com/introducing-girls-to-code-one-flower-at-a-time/ . [19] Kendra Wannamaker, Wesley J. Willett, Lora A. Oehlberg, and Sheelagh Carpendale. 2019. Data embroidery: Exploring alternative mediums for personal physicalization. Conference poster, May 2019, University of Calgary, Calgary, Alberta, Canada. http://hdl.handle.net/1880/110218 [20] Jordan Cunliffe. 2022. Record, Map and Capture in Textile Art: Data Visualization In Cloth And Stitch. Batsford, London, England. https://www.rizzoliusa.com/book/9781849947190/ [21] Utica University. Bloom’s taxonomy of measurable verbs. Retrieved September 15, 2025 from https://www.utica.edu/academic/Assessment/new/Blooms%20Taxonomy%20-%20Best.pdf . [22] Naomi Milgrom Foundation. 2023. MPavilion 10 designed by Tadao Ando. Retrieved from https://mpavilion.org/design-tadao-ando/ . [23] City of Melbourne Open Data Team. 2023. City of Melbourne Open Data. Trees, with species and dimensions (Urban Forest). Retrieved from http://melbourneurbanforestvisual.com.au/ . [24] Cassandra Hradil. 2023. Embodied information: Local trees, datasets, and you. Workshop spring and fall 2022. University of Pennsylvania, Philadelphia, Pennsylvania, USA. Retrieved from https://cassandrahradil.com/projects/woodlands/ . [25] Stefanie Posavec. 2025. Analog data visualization for storytelling. Online course accessed via Domestika.org. https://www.domestika.org/en/courses/3964-analog-data-visualization-for-storytelling/course

Appendix a

1.INTRODUCTION

1.INTRODUCTION

Booklet for “Data Stitching and Storytelling” workshop, February 2024. Worksheets within were designed to help participants record observational data about each tree, translate that data into visual variables, and sketch their visualisation before stitching.

Figure A1: Booklet pages 1-2. The cover (left) and workshop agenda (right).

Figure A2: Booklet pages 3-4. A summary of the workshop objectives and the five steps of data visualisation (left); and a map of trees (numbered and highlighted) for observation (right).

Figure A3: Booklet pages 5-6. An ideation worksheet to brainstorm questions for data collection, measurements to answer those questions, and ways to visually represent those measurements (left); and a reference sheet containing a table of tree data (map number, name, and diameter) from the City of Melbourne’s Open Data dataset (right)

Figure A4: Booklet pages 7-8. Field data collection sheet for participants to record measurements or observations for up to three variables (left); and diagrams of the three embroidery templates available for participants to stitch (right).

Figure A5: Booklet pages 9-10. An encoding worksheet to help participants connect tree variables to visual attributes (left); and a stitching legend worksheet where participants could either sketch how they would stitch each variable or tape a sample of the material (thread or bead) they would use to encode each variable (right).

Figure A6: Booklet pages 11-12. A diagram of how to thread a needle, tie a knot, and sew basic stitch patterns (left); and a page with additional workshop information, including resources and references, a QR code to a post-workshop survey, and links to connect with The Data Garden Co. on the web and through social media platforms.

Appendix B

1.INTRODUCTION

1.INTRODUCTION

Participants could sketch their embroideries on one of the three paper templates before stitching.

Figure B1: The sketching worksheet for participants who chose the dot grid template. 

Figure B2: The sketching worksheet for participants who chose the donut shape template.

Figure B3: The sketching worksheet for participants who chose the map template.

TABLE OF CONTENTS

TABLE OF CONTENTS

TABLE OF CONTENTS

ABSTRACT

ABSTRACT

Introduction

Introduction

INTRODUCTION

Background

Background

BACKGROUND

Objectives

Objectives

OBJECTIVES

planning and logistics

planning and logistics

PLANNING AND LOGISTICS

MATERIALS

MATERIALS

MATERIALS

WORKSHOP EXECUTION

WORKSHOP EXECUTION

WORKSHOP EXECUTION

WORKSHOP outcomes

WORKSHOP outcomes

WORKSHOP OUTCOMES

Suggestion for future workshops

Suggestion for future workshops

SUGGESTION FOR FUTURE WORKSHOPS

Conclusion

Conclusion

CONCLUSION

acknowledgements

acknowledgements

ACKNOWLEDGEMENT

REFERENCES

REFERENCES

REFERENCES

aPPENdix a

aPPENdix a

APPENDIX A

APPENDIX b

APPENDIX b

APPENDIX b