Data Visualization & Display Techniques in BI

Why is data visualization and proper display techniques critical in a business intelligence environment? What are some of the potential pitfalls to avoid when presenting analytics for end users?

Data visualization and proper display techniques are important to improve transparency of data. Clear and appropriate use of graphs and other resources to display data help in facilitating a clear understanding of data trends to aid in end user decision making. Strome (2013), states “data visualization is more than simply an alternative to presenting data in a table,” in fact, it’s used to tell stories, illustrate problems, and demonstrate trends. Analysts who prepare dashboards or statistical process control (SPC) charts must design and provide the information in a way that’s relevant to organizational goals. They must also avoid information clusters and avoid information “push” as this can cause people to become passive recipients of information (Strome, 2013).

Preparing data to be presented to leadership or other members of the organization should be legible and understandable. Park, Smaltz, Mcfadden, & Souba (2009), argue that a dashboard cannot merely contain data, but needs to be presented with specific goals in mind. Proper display of information and providing related indicators can help an organization function appropriately. Reflecting on the success of St. Vincent Health System (SCHS) through use of different model dashboards and trial and error processes, they were able to develop a dashboard that helped improve nursing quality initiatives and shared governance structure within the organization (Frazier & Williams, 2012). Taking the focus from quantity and focusing on quality, the organization learned that this would help declutter dashboards and make it more understandable for those who used them.

One of the major “pitfalls” of data presentation is information “push”. Understanding your recipient’s data needs and goals for the group or individuals are key. In many cases presenting data to individuals who only require bi-weekly or monthly updates may become passive recipients if information is presented daily of weekly. One of the major issues that I see in my organization is the “push” we get daily of information that is irrelevant. The issue with this is that it simply goes in my trash bin because I have no use for it. Strome indicates another issue is that we become reliant on this information and fail to query the data for our own research. To improve overloads one recommendation is alerts that are sent via email to indicate important changes. This prompts recipients to research the data and act on any changes that need to be made (Strome, 2013). 

References:

Frazier, J., & Williams, B. (2012). Successful implementation and evolution of unit-based nursing dashboards. Nurse Leader, 10(4), pp. 44-46. http://doi.org.proxy.cc.uic.edu/10.1016/j.mnl.2012.01.003

Park, K., Smaltz, D., Mcfadden, D., & Souba, W. (2010). The operating room dashboard. Journal of  Surgical Research, 164(2), pp. 294-300. http://doi.org.proxy.cc.uic.edu/10.1016/j.jss.2009.09.011

Strome, T. (2013). Usability and presentation of information. Healthcare Analytics for Quality and Performance Improvement. Pp 165-181. Hobroken, NJ: John Wiley & Sons, inc.