Data visualization represents data and information in graphical forms to help decision-makers understand and analyze complex information faster. Data visualization takes many forms, such as graphs, charts, maps, and other graphic aids (Wang, Sundin, Murray-Rust & Bach, 2020).
The importance of data visualization:
The importance of data visualization lies more in quick decision-making, as in the business environment, especially when working in construction, data must be interpreted quickly and accurately. Data visualization provides seeing patterns and relationships between data that may not appear in regular tables at first glance and helps interpret the final results of this data. It also identifies outliers outside the expected range of values and is critical in identifying potential problems. Finally, it has been proven that individuals than complex data better remember images (Post, Nielson & Bonneau, 2002).
Data visualization bias:
According to Gaslowitz, L. (2017), one biased approach is to manipulate the axis scales by selecting a specific range of axis values and presenting them as evidence that the difference between the two points is more significant or more minor than it actually is. You find this in the differences between structural systems' durability and lifespan. Another way is to reverse the direction of the axis to make it appear as if the line is heading in the opposite direction. Also, selective data points can be used to use it as if it is all the data present, or even use misleading captions or have two meanings. There is also the use of a pie chart to represent and stack many categories, thus making it difficult to compare them.
Avoiding bias in data visualization:
In the field of construction, we, as engineers, must represent the data with very high accuracy, as data manipulation may lead to loss of life when buildings are completed and made available to the public. This can be done by reviewing the results through a more significant number of sectors concerned with construction projects (Epstein, 2020). For example, we deal with several sectors of government projects, such as contract management, procurement, and technical departments, such as architectural, construction, electrical, and mechanical.
It is good to specify a specific question that must be answered through data visualization and not leave room for general information with more than one meaning. Detailed information is required about the data selection process and how it was analyzed, and then the inaccurate ones are excluded. Using the appropriate data visualization in the right place gives a better and more reliable impression that the data provided is not tampered with (Szafir, 2018).
Conclusion
Data visualization constitutes a more significant interaction of people as it can be understood quickly and facilitate complex information. It is also an essential tool for interpreting results for decision-makers. To avoid bias in data visualization, researchers must ensure that it has not been manipulated to support a particular conclusion. Transparency regarding the selection of data visualization is essential to avoid bias and help readers to access complete and entirely correct information.
References
Epstein, Michael (2020). Data Visualization and Misleading Representation of Information. Retrieved from:
http://www.amplaboratory.org/classes/statistics/statistics-unit-4-data-visualization/ (pass: teaching stats).
Gaslowitz, Lea. (2017). How to spot a misleading graph [Video file]. Retrieved from https://www.youtube.com/watch?v=E91bGT9BjYk (4:09)
Post, F. H., Nielson, G., & Bonneau, G. P. (Eds.). (2002). Data visualization: The state of the art. https://www.google.com/books/edition/Data_Visualization/WAZYsfeMi4kChl=en&gbpv=1&dq=data+visualization+techniques&pg=PR9&printsec=frontcover
Szafir, D. A. (2018). The good, the bad, and the biased: Five ways visualizations can mislead (and how to fix them). interactions, 25(4), 26-33.
Wang, Z., Sundin, L., Murray-Rust, D., & Bach, B. (2020, April). Cheat sheets for data visualization techniques. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems (pp. 1-13).