Can The Public Trust The Numbers If They Can’t Understand Them? ; A Research Paper Summarization on Covid 19 Vaccination Data Visualization Issues

M. Shabazz — Mon, 12 Oct 2026 03:27:48 +0000

Main Research Questions and Why They Matter
The main research questions asked by the research paper authors were: What are issues with visualizing vaccination data for Covid 19? And how can Covid 19 Data be better visualized? This information matters because Covid 19 was declared a pandemic by the World Health Organization and data is only useful if it can be interpreted correctly. The data can have all the information needed but if the reader is unable to consume the data in a manner that it is understandable then decisions based on the data may be flawed.

Why Main Research Questions are relevant or Important

During covid there were many dashboards, charts, graphs, etc presented to academia, healthcare professionals and the general public. As a software engineer and an ER nurse I was able to see both sides of the issue. The engineers and data analysts creating the dashboards and graphs with their understanding and the healthcare professionals attempting to make sense of the graphs and numbers, and large quantities of data pouring in. A chart that is not easily understandable or appears confusing can affect trust in the data and can skew critical decisions such as whether people decide to vaccinate themselves or not. Accessibility is also a consideration because Covid 19 charts often had colorful data but care needed to be taken to ensure people who were colorblind were not left out of being able to decipher critical data.

Fig 1. Fig 6 (Ismail et al., 2022, p.305) from the research paper shows a non effective dashboard, it lacked interactivity and thus was not as useful.

Outline of Research Methods Used

The researchers built several dashboards to show vaccination progress, effectiveness and acceptance (Ismail et al., 2022, p.301). The two data sources used included Covid 19 pandemic dataset from the Ministry of Health Malaysia. This data set was made available to the public via Github (https://github.com/MoH-Malaysia/covid19-public), (Ismail et al., 2022, p.305). The data set looked at daily ICU admits and vaccinations for Covid 19 in Malaysia. The researchers also used a data set taken from an online survey given to people in Malaysia asking why they had not gotten a Covid 19 booster and 98 people responded (Ismail et al., 2022, p.305). The dashboards were then reviewed using visualization literature for decisions on design and compared with known successful dashboards.

Key Findings

Some of the charts used color without any clear purpose. For example two charts (Fig.2 & Fig 3) have every bar a different color in a rainbow spectrum but there is no clear purpose for this color range and it does not enhance data visualization or viewer understanding. Also the color can be hard to see for some people with color blindness.

Fig 2. A rainbow bar chart (Ismail et al., 2022, p.303), where the color spectrum serves no clear purpose and is not accessible.

Fig 3. A rainbow bar chart (Ismail et al., 2022, p.303), where the color spectrum serves no clear purpose and is not accessible.

As mentioned above (Fig 1) some charts lacked interactivity. Which did not allow users to extract key data that was individual to their needs. Being able to filter, zoom, etc is needed to find data that matters most to a viewer.

Having too much data in a chart can also present an issue. An example is Fig 4, where the vaccination progress for each state and each age group is shown in one chart. A fix was separating out the age groups to use buttons to filter between them which then allowed users to easily extract the information they wanted.

Fig 4. This chart is filled with data and that is part of the issue, because there is so much data in one chart it can be hard to find the specific parts a viewer is interested in. (Ismail et al., 2022, p.305)
.

Limitations of Research

One limitation was the survey used only had 98 responses and doesn’t discuss how survey responses were explicitly gathered. Also they cited many blogs as sources rather than peer reviewed research: for example ‘storytelling with data’, ‘How to Create and Customize a Treemap Chart in Microsoft Excel’, etc.

The researchers also mentioned using grey in one chart (Ismail et al., 2022, p.306), to help color blind viewers understand the chart , however they never tested whether the chart was viewed better by an actual color blind viewer.

My Conclusions and Takeaways

My main takeaway is that first consider the story you want your chart to tell or what data you want to present to the viewer. Then design your visualization around the purpose and story of your data visualization. Also I found the issue of too much data at once as something I can easily connect to from a clinical standpoint. In the ER when too many items are flagged in the red , alarm fatigue happens visually. What is truly critical is lost in the sea of items that may be important but don’t all have the same level of priority red should be used to indicate. Good visualizations don’t just ask the user to trust the numbers, they present the numbers in a way that earns the viewers trust.

Questions to Consider

What kind of studies needed to be used / designed to test whether the researchers’ grey chart actually was helpful for color blind people?
Filters can be used to sort through useful data but they can also hide data if the initial filter is too narrow? How do you strike a balance?
How can we reduce visual alarm fatigue in healthcare professionals in an age where large amounts of data are consistently available? How do we prioritize safe healthcare while not overloading our healthcare professionals?
What accessibility issues do public health dashboards have and should all public health data be required by law to be accessible to anyone with any accessibility issue?

References