Beyond Dashboards: Supporting Cognitive Work Through Deeper Problem Understanding

Posted by Lisa Douglas, PhD on December 21, 2020
Beyond Dashboards: Supporting Cognitive Work Through Deeper Problem Understanding

An effective decision support tool requires deep domain knowledge, understanding of the problem space, and consideration of multiple perspectives while capturing the data relationships that really matter.

Table of Contents

Dashboards—touted as “at-a-glance” decision-support tools that display performance indicators or metrics—sit atop many companies’ current wish lists. These tools convey a promise to help users know more, make decisions faster, better understand organizational status, and meet business goals. Ideally, a dashboard takes complex data sets and transforms them into visualizations that quickly provide a deep level of understanding and an ability to uncover insights about their business portfolios.

Mile 2 understands the attraction to distill down complex systems and complex information into simple visualizations. The idea of rapid decision making while navigating a complex problem space is advantageous in almost any professional environment. But oversimplification can be a problem when the simplification breaks the meaningful connections and interactions within a system. The model of the dashboard itself fails to capture the significant dimensions of the work and decision space. Dashboard solutions, whether off the shelf or custom-built, are often marketed as easy to implement and simple to use. However, the challenge lies in representing complex data in only one view, when that data can impact an enormous amount of people and their cognitive work.

Understanding Dashboard Implementation Challenges

During the design and development phase of a dashboard, or any decision-support tool, a few important questions need to be answered:

  • What do the creators understand about the domain and the problem space?
  • What is the true function of the dashboard?
  • What is the relationship between the different types of data, if any, and how will this impact decision-making?

Decision-support tools are often marketed with the appealing ideas of faster, better, more efficient business operations, but these goals can be meaningless without a deep understanding of the domain and problem space where speed, accuracy, and efficiency challenges arise to begin with. If a primary user goal for wanting a dashboard is efficiency, then the creators need to understand the most critical and common issues when implementing a dashboard, such as data obfuscation. Suppose a dashboard user recognizes that the information frequently used to make important decisions is located in several disparate systems. Bringing all the disparate data into one visualization sounds like a time-saver. In reality, the presentation of the data often has to be reinterpreted because the real data are obscured in the graphics or visualizations that were intended to improve efficiency.

Mile 2 uses a unique approach to understand the pain points and constraints involved with a user’s business decisions that will highlight the need for better presentations up front. For example, the simpler visualization may not represent the data accurately or holistically. Skewed scales on graphs might obscure the relative magnitude difference between categories on a bar chart—they will look very different on a small range versus the actual range of the data. Working collaboratively through these issues during the creation and development phases ensure decisions are not made on degraded information.

Function Over Form

Another challenge that surfaces when implementing a dashboard is graphics and visualizations are often selected for their design and not their function. Visualizations that do not properly reflect the underlying cognitive work can result in users trying to understand the graphic instead of the actual problem. A good example of this phenomenon is the use of a pie chart. The purpose of a pie chart is to show proportions, but this chart type has to be large enough for its data to be clearly visible. Usually, a pie chart will be presented larger than other elements in the visualization, which may complicate a user’s understanding. Conversely, treemaps can help users with proportional data in a more compact way and are easier to compare side-by-side if additional data comparisons are needed.

At Mile 2, we use a variety of discovery and understanding techniques and methodologies to define what information our clients need to improve their decision-making processes. Our scientists and designers have extensive experience in data representation and understand how to mitigate confounding information. We work directly with clients to design the right visualization with the right amount of fidelity.

Relationships Put Data into Context

Dashboards introduce significant risk to understanding relationships between the visual components as well. Decision-making in complex socio-technical systems is based on understanding the relationships between a lot of moving parts, and typical dashboards often fail to visually coordinate multiple frames of reference. When the dashboard does not explicitly connect these frames of reference, they are left as disparate frames visualized on the dashboard. The user then has to interpret and integrate these frames, negating the intended goals of dashboard implementation.

Mile 2 uses representation aiding techniques like exploring multiple frames of reference; putting data into context; highlighting contrasts, change, and pivotal events to provide a more holistic view that users need to make sense of the data. Displaying change over time is a critical and powerful technique to show temporally significant differences as the world (and our model of the world) constantly evolves.

Turning Challenges into Tailored Solutions

At Mile 2, our clients are integrated into the design and development process along with us. Through our cognitive systems engineering (CSE) approach, they gain insights about their own work processes through our elicitation and observation techniques. We consider all applicable patterns of cognitive work that can be generalized, such as attention redirection, intelligence analysis, anomaly response, medical diagnosis, data interpretation, and contingency planning or re-planning. Understanding cognitive work means also understanding joint cognitive systems, which simply refers to systems composed of people and machines. For our diverse clientele, machines could mean algorithms, sensors, robots, and other digital tools.

Mile 2 tackles hard problems in complex socio-technical systems—resulting in our customers gaining a trusted partner in the development of decision-support tools.

Connect With Us

We bring our expertise to interpret and translate complex challenges into technology that can give your organization a competitive advantage in your field. Start the conversation with us and discover how Mile 2 can help solve your toughest challenges.

Copyright © 2024 Mile 2, LLC.
All rights reserved. Web design by Jetpack