Creating a DIKUW chart

Last updated Sep 25, 2024 | Originally published Sep 25, 2024

As described in Data, Information, Knowledge, Understanding, and Wisdom, we make data from phenomena we observe in the world. When we add meaning to data, we gain information, which helps us make decisions and take actions. When we put information into practice and learn from the process, we gain knowledge. When we combine knowledge with other knowledge — especially knowledge about knowledge, such as knowledge about whether the quality of knowledge we have is good, or knowledge about how to apply knowledge — we gain understanding. Each of these levels help us with efficiency (Ackoff, 1989): we use them to attain a specific goal, and as we gain data, information, knowledge, and understanding about that goal, we get better at attaining it. That is, we learn to attain that goal more predictably and while consuming fewer resources each time. However, it is no use to be efficient if we are efficient about doing the wrong thing. To be effective, according to Ackoff (1989, is to do the right thing. Effectiveness is about attaining the most valuable goals. Data, information, knowledge, and understanding can only help us be more efficient in the attainment of goals — but to judge whether those goals are effective, we need wisdom. Design is the art and science of designating importance: of finding and framing the right problems, determining the criteria for solving those problems, and solving them by finding solutions that best fit those criteria (Simon, 1995). Wisdom is therefore produced by applying design to understandings. We do that by asking (and seeking the answers to) questions like “What problems are most important?” “Which gaps in our knowledge are critical?” “What do we not understand or know?” “Where are we wasting efficiency on the wrong goals?” Questions like these form the roots and foundations of the theories we use to explain and predict the world and prescribe the most valuable goals and how to attain them efficiently.

Okay: so what? How do we use these ideas about data, information, knowledge, understanding, and wisdom (DIKUW) to better ourselves and our organizations? A DIKUW chart is a simple way to put these ideas into practice. DIKUW charts combine Sanders’s (2015) sensemaking framework with Basadur’s challenge mapping (Basadur et al., 2000). DIKUW charts connect our key, critical questions with what we know and what we don’t know — and how we know, or how we can learn, those things. The goal of a DIKUW chart is to model our wisdom, understanding, knowledge, information, data, and the phenomena that we derive them from. Creating a DIKUW chart is therefore a way of practicing wisdom: of reflecting on what matters and what doesn’t, identifying the important gaps, and charting a path to resolving those gaps.

Figure 1 provides a demonstrative sample of a DIKUW chart. This sample DIKUW was developed to inform a research project on data crowdsourcing. At the top of the chart are the theories we have (theory of classification, theory on conceptual modelling) and the theories we are trying to build (“how to design for crowdsourcing contributions in data crowdsourcing.”) It then breaks those theories down into components, all the way down to the phenomena that make up the problems and solutions we are concerned with.

The process of developing a DIKUW chart is simple. However, it is also iterative. As you go through this process, work in pencil, or marker and sticky notes. Be willing to revise, move, remove, and add to the chart as it develops.

Begin with a goal. What are you trying to achieve? Remember that while wisdom is about deciding what the best goal is, there is no perfect answer, here. Design is tentative and iterative. Start somewhere — anywhere — and give yourself permission to change later. In figure 1, the goal is designing for [better] crowdsourcing contributions in data crowdsourcing projects. Write that goal at the top of a chart.

Then, ask: what understandings do you need to achieve that goal as efficiently as possible? Write these understandings down as separate elements on the chart and connect them to the goal you’ve identified. In our example, a key issue in crowdsourcing is the variety of contributors that might contribute to a given crowdsourcing project. Different contributors may have different skillsets or expertise or background that influence the qualities of data they are able to share with the project. So, the understanding highlighted in figure 1 is that crowdsourcing projects should account for this variance.

Third, what knowledge can we use to inform that understanding? Connect these pieces of knowledge to the understandings you added previously. In the example above, we know that the interfaces people use to contribute to crowdsourcing projects come from the conceptual model of the project: how the project designers see the world, who they expect contributors to be, and how they expect them to be able to contribute. They design the contribution interface according to that conceptual model, so potential contributors who do not match the conceptual model of the project designers might run into friction when they try to contribute (e.g., they might not speak the same language as the interface, leading to mistakes). However, we don’t know what influences contributor motivation. This is where the DIKUW starts to become useful: by identifying what we don’t know, we learn what we need to learn.

Fourth, what information provides the knowledge we just noted? In the example, this takes the form of the conclusions of scholarly studies — but this need not be the case. Information is simply data with meaning. Any input you can think of that confirms your knowledge is worth adding to the chart. So, add the pieces of information you can think of that contribute to the knowledge you identified before and connect them in the chart. At the same time think about what information would provide the knowledge you are looking for. Remember that the purpose of a DIKUW chart is not only to map what you know, but also to map what you need to figure out. Add these ideas and connect them as well.

Fifth, what data drives that information? Label these on the map and connect them accordingly. In the example, experiments and surveys informed the scholarly studies we cited. There are many other kinds of data that could be useful, however. Reports, dashboards, or metrics in your organization, for instance, provide valuable sources of data that may be transformed into information in service of your developing theory.

Sixth, what are the phenomena in the world that we turn into those data? Consumer behaviour? Products and services? Staff functions and roles? Anything goes, as long as it can be observed and turned into data. Add these phenomena.

As you go through these steps, feel free to re-work the map. Does something you added as understanding actually fit better as knowledge? Move it! Are you thinking of new key questions as you work? Add them! Does something you’ve just added map to multiple parts of the chart? Sketch the relationships out. Did you just think of a new piece of data or information that might be useful, but you’re not sure how yet? Put it down and see what comes of it.

The purpose of the chart is not actually the chart itself, but what you learn from the process. The goal of engaging in a DIKUW charting exercise is to reflect and reflect on your learning or your organization’s learning. The chart is a working model of what you know, how you know it, what you need to know, and how you can figure it out. So, remember that all models are wrong (Box, 1976) — but it is your responsibility to make sure that this model is useful.

Ackoff, R. (1989). From data to wisdom. Journal of Applied Systems Analysis, 16, 3–9. http://www-public.imtbs-tsp.eu/~gibson/Teaching/Teaching-ReadingMaterial/Ackoff89.pdf

Basadur, M., Potworowski, J. A., Pollice, N., & Fedorowicz, J. (2000). Increasing understanding of technology management through challenge mapping. Creativity and Innovation Management, 9(4), 245–258. https://doi.org/10.1111/1467-8691.00198

Box, G. E. P. (1976). Science and statistics. Journal of the American Statistical Association, 71(356), 791–799. https://doi.org/10.2307/2286841

Murphy, R. J. A., & Parsons, J. (2020). What the crowd sources: A protocol for a contribution-centred systematic literature review of data crowdsourcing research. AMCIS 2020 Proceedings, 20, 6. https://core.ac.uk/download/pdf/326836069.pdf

Sanders, E. B.-N. (2015). The fabric of design wisdom. Current, 06. https://current.ecuad.ca/the-fabric-of-design-wisdom

Simon, H. A. (1995). Problem forming, problem finding and problem solving in design. Design & Systems, 245–257. http://digitalcollections.library.cmu.edu/awweb/awarchive?type=file&item=34208