How do we apply visualization guidelines when reality knocks on the door?

[Hyoideum]

When theory meets reality – thoughts by a student in visualisation

Like many I had a gap year after high school. During this time, I worked at different companies in production. I remember looking at some of the monitors, with beautiful visual charts, in the middle of a busy shift thinking: “Why does it look so complicated, as if we have time to spare decoding rather simple information?”

Now I am sitting on “the other side of the table”, inexperienced in visualisation, wondering.

How do I navigate implementation, when theory meets reality? How do I avoid the pitfall of focussing so much on the many useful guidelines and theories that I forget the bigger picture?

What might fit in one situation of a workday, may be unhelpful in another. On top of that, often multiple users with different needs may use the same visual charts.

We learn about theories such as Munzners framework, but without direct contact with the end user can we truly apply the WHY confidently?

Let us do a thought experiment.
We work on visualising information that is used by workers handling a production line. Different products will be manufactured here. At 09.00 o’clock we produce product A. It is very important to monitor the heat properly otherwise the production experiences repeated temporary stoppages costing the company money. At 12.30 we change product. This product needs monitoring the speed rather than the heat. At 16.00 another product with another important feature is produced.
The change in environment has suddenly changed the important information and what might warrant a preattentive feature in the visualisation. Now if we do not change the feature, the user may be guided to look for the wrong parameter. We could change the charts depending on the product I suppose, but then the staff needs to familiarise themselves with multiple different charts which may increase the effort to interpret the visualisation.

In the paper “Design Study Methodology: Reflections from the Trenches and the Stacks” Sedlmair et al. (2012), identify a challenge in getting to know the user and designing visualization that meet their needs even when doing a dedicated design study. I am wondering how we address this in everyday visualization work, where the designers may work at a different location to the users, making observing their working condition rather difficult.

I believe what ultimately determines the proper choice of visuals would be the individual context of its user rather than the guidelines themselves. I do believe guidelines are truly helpful and necessary to make sure that the quality of visualisation is kept up to the best of standards. I am worried though, that a combination of not knowing the user well enough and a too narrow view on how to apply visualisation guidelines, may cause me to lose sight of why information is visualized in the first place.

How do designers maintain that broader perspective when guidelines are specific, sometimes potentially conflicting while time and access to end users are also limited?

[alvitta]

You raise several questions that many of us working in the VIS community grapple with. The relationship between theory, guidelines, and practical design is something I’ve been thinking about through my own research, so I thought I’d share a few thoughts.

On the relationship between theory and practice

I agree that translating visualization theory into practice can be difficult. Theories and frameworks typically consolidate knowledge from various sources to identify general patterns and offer ways to think about problems. For example, Munzner’s framework draws on existing visualization research to describe different levels of visualization design, from understanding the problem to implementing a solution [1].

One challenge with these frameworks is that they necessarily abstract away some of the details that make each situation unique. They help us organize our thinking and identify questions worth asking, but applying that knowledge to a particular workplace, task, or group of users still requires understanding the local context. I think that’s part of the challenge you’re describing.

On the role of visualization guidelines

My research group has been studying visualization guidelines, including interviewing people who develop them, and some of what we’ve learned might be relevant to your questions.

One observation is that organizations often create guidelines to address specific needs [3]. For example, a company might want consistency across its visualizations or need a way to communicate design expectations to external consultants. In these situations, guidelines serve as a shared reference for how to design visualizations.

We’ve also found that guidelines aren’t necessarily grounded in visualization research. Sometimes they reflect the experiences and preferences of the people writing them. One author we interviewed, for example, recommended donut charts over pie charts simply because they preferred donut charts.

We also don’t know much about how widely these guidelines are used. Our research focused on publicly available guidelines, and we don’t know what proportion of organizations have formal guidelines at all. I suspect that many visualization decisions are made through designers’ own experience, knowledge, and preferences.

Your question about conflicting guidelines also connects to something we’ve been exploring. Across organizations, recommendations sometimes differ, which makes sense given that each organization has different priorities. Within a single organization, guidelines often include conditions or exceptions that help explain when a recommendation applies [2].

Results from perceptual studies are similar. For instance, two studies comparing bar charts and pie charts might have different conclusions because they examine different tasks or participant groups. When we summarize those findings into simple recommendations, we lose some of that context. This is something I think we could improve in both research and practice: communicating not only what is recommended, but also why and under what conditions.

Thinking through the production-line example

Your production-line example is a useful way to think about how information needs change throughout the day. It also made me wonder what the workers are actually trying to accomplish with the visualization.

For instance, temperature and speed are both quantitative variables. Depending on the task, the same chart design could work for both, with the displayed variable changing by product. That might let the interface adapt while keeping a familiar design.

But there are other possibilities too. Are workers looking for trends, monitoring thresholds, or identifying unusual changes? Do they need to interpret a chart, or would an alert or status indicator be more useful? I can imagine different solutions depending on the answers.

Without knowing more about the working environment and the decisions workers need to make, it’s difficult to know which approach would work best. And I think that connects directly to your broader question about understanding users.

Designing when we don’t have all the information

Your point about limited access to users is one of the practical challenges of visualization design. Ideally, we would have opportunities to observe people in their working environment and understand their tasks before designing anything. But that isn’t always possible.

In those situations, we might learn from domain experts, existing documentation, or the tools people already use. We might also develop a few alternatives and gather feedback when possible. Of course, the options available will depend on the time, resources, and access we have.

I think your emphasis on context is a useful way to approach visualization design. Theories and guidelines give us knowledge to draw on, but we still have to decide how that knowledge applies to the situation we’re working with. And sometimes understanding the problem leads us to reconsider whether a visualization is the best solution at all.

Papers referenced

[1] Munzner (2009). A Nested Model for Visualization Design and Validation
A framework for reasoning about visualization design and evaluation.

[2] Ottley (2026). Consensus and Contradictions: A Cross-Organizational Analysis of Visualization Style Guides An analysis of publicly available visualization guidelines, examining shared recommendations and differences across organizations.

[3] Ottley & Schwabish (2026). Data Visualization Style Guides in Practice: Why They Emerge, How They Work, and When They Bend Interviews with guideline authors about why organizations create guidelines, how they use them, and when exceptions are permitted.