Thoughts and Inputs about the new role of dashboards in decision-making
The era of dashboards has reached us as digitalization and knowledge management professionals. What will the impact be? I am sending a few thoughts about the impact and innovations. Let’s put terms in context first:
Globalized dashboards have become feasible because powerful tools like e.g. ArcGIS, PowerBi, R statistical programming language are relatively new players in the analytics arena and might support wisdom which impacts decision-making processes faster than ever. The new tools and big data analytics as a field require new competencies from digitalization professionals and knowledge managers especially, in order to be applied meaningfully (e.g. a subject related solid foundation in statistics).
Dashboards will play an important role in societies of the knowledge age. There will be tremendous opportunities at a global scale as well as globalized fatal risks, which we have not seen before - comparable to e.g. atomic energy, road traffic, chemical industry from the industrial age. Dashboards (or cockpits) have changed the high-tec industry for decades already, just have a look at the navigation and control interfaces of Apollo, Discovery, Dragon V2 . They have been used to simplify the control of complex machines (hardware and software).
The dashboard building blocks: the Data Collection, the Quantitative Data Model, the Visualization
Let’s map the terms wisdom, knowledge & information, data on 3 levels of a dashboard design process. I am going to present a series of questions regarding a dashboard quality whereas each has similar, critical relevance.
The visualization
The visualization, what we see as users functions as wisdom and as a knowledge dissemination layer. Some guiding questions for the design:
The Quantitative Data Model
The Quantitative Data Model applied comes with in-built knowledge concepts valid in specific fields. It is used to process and aggregate the data for the visualization layer. Some guiding questions for the design of a dashboard:
The choice of the quantitative model has a tremendous impact in today’s networked world. I am illustrating with an example: Let’s say you measure the existence of a virus (e.g. the herpes zoster virus / shingles), which is dangerous. A quantitative data model used by a medical lab may need only simple quantitative data like e.g. the contamination of a population sample in a region. Contamination could reach figures as high as 30% in case of the herpes zoster virus. How relevant is this data generated from the quantitative data model for other decision-makers who do not work e.g. in a lab? A quantitative data model used by epidemiologists or medical diagnostic experts must allow for the aggregation/relation of e.g. contagious (which differs from contamination) and causal ill patients in the population sample and e.g. the probability of transmission for those who have not been exposed. The data output needed by 2 different professional groups working even in a related field can vary by powers. Dependent on the choice of the quantitative data model, visualization and subsequent decision-making can differ tremendously.
The correct choice of the quantitative data model can lead to unknown blessings on a global scale whereas an inadequate choice can lead to unknown worst-case scenarios, too.
Recommended by LinkedIn
A very recommendable reading from Sarun Charumilind about quantitative models: Demystifying modeling: How quantitative models can—and can’t—explain the world
The data generation
Data generated must comply with the 3 main quality criteria. Any violation of these mandatory criteria would lead to significantly false dashboard visualizations (the GIGO principle – garbage in, garbage out).
So, what's new?
Big data analytics and dashboards are technology as well as social game-changers. Why social game changers?
Dashboards have been serving as interfaces to simplify the control of sophisticated technology systems in the high-tech arena since decades (see the link to the Dragon V2 dashboard above). Now, they are massively entering other areas of decision-making in complex social systems (e.g. public health and project monitoring, in general). Thus, in contrast to the deployment for command and simplified control in the technology arena, there will be complex social dimensions (like e.g. impact on the quality of life) to be considered when building a dashboard with a social impact. Some dimensions are hard to measure quantitatively for visualization as live charts because they require qualitative measurement and human intervention, e.g. the profound interpretation of charts and their social impact.
Especially, knowledge managers who design and build dashboards, get into much more powerful positions. A new ethical role as trustworthy sources emerges. They prepare e.g. policy decision-making more than ever because with the advent of the powerful statistical web-based software raw data is visualized as common knowledge and pretends to be wisdom instantly - the visualization on dashboards pretends to be unquestionable truth in the mind of the knowledge consumer. Bostrum (2011) relates to the consequences of true information, elaborates on the term Information hazard, and offers a taxonomy. Information hazard is a risk that arises from the dissemination of true information (e.g. the figures and charts from test results) that may cause harm or enable some agent to cause harm.
We have known live-dashboards from the stock market, now they are becoming universal tools for live visualization of complex underlying social phenomena. We all know the John Hopkins U Dashboard which had billions of page views on single days and it was directly used for social and policy decision-making in most national crisis management centers on this planet.
I see a paradigm shift happening in the field of knowledge management especially. Ubiquitously, globally accessible dashboards pretend to mirror a globalized, homogenous environment. In the reality of the people there can be little social similarity e.g. if we put the people living in the city center in Lilongwe in Malawi and a small village in Germany on the same dashboard.
So, because of the high risks at each level in the dashboard design process and the limitations of quantitative data models, qualitative measurement (e.g. impact assessment), as well as profound interpretations of quantitative visualizations, must go hand in hand when using dashboards in social decision-making processes.
Are the dashboard producers ready for this challenge and their new responsibilities? Are the consumers (e.g. policy decision-makers, society, professionals) ready? What about critical thinking in the era of dashboards, which now requires a deep understanding of the black box, the invisible, underlying quantitative data model applied?
References and Further Reading:
