What makes a great data story?

A data story is a narrative that explains the insights and meaning behind a set of data or a data analysis. It is a way to communicate the key findings and conclusions of a data analysis in a clear and compelling way to a broader audience. A data story combines the power of data visualization and storytelling to create a compelling narrative that engages the audience and helps them to understand the insights and implications of the data.

A data story typically follows a structure that includes:

1. Introduction: This section sets the context and introduces the data and the problem that the analysis is trying to solve.
2. Data exploration: This section presents the data and highlights any interesting patterns or insights that emerge from the data exploration process.
3. Data analysis: This section explains the analytical methods used to analyze the data and presents the key findings of the analysis.
4. Interpretation: This section interprets the findings of the analysis and explains their implications and significance.
5. Conclusion: This section summarizes the main findings and conclusions of the analysis and provides recommendations for future actions or research.

A good data story is clear, concise, and engaging, and uses data visualization to communicate complex information in a simple and intuitive way. It should also be tailored to the needs and interests of the target audience, and should be designed to be easily understood and remembered.

The great story: “How the Virus Won”

One great data story from 2020 was the New York Times article “How the Virus Won,” which used data visualization to tell the story of how the COVID-19 pandemic spread across the United States.

The article used a combination of interactive maps, charts, and animations to show how the virus spread from its initial outbreak in the Northeast to the rest of the country, and how different states and regions were affected at different times. It also showed how policy decisions and public behavior affected the course of the pandemic, and how the country’s response to the virus evolved over time.

The data story was a great example of how data visualization and storytelling can be used to make complex information accessible and engaging to a broad audience. It provided valuable insights into the spread of the virus and the factors that contributed to its impact, and helped to inform public understanding and policy decisions around the pandemic.

The bad story: politically manipulated

One bad data story from 2020 was the controversy over the COVID-19 case reporting in Florida. The state government was accused of manipulating data and underreporting the number of cases to make it appear as if the pandemic was under control.

The story was a bad example of data reporting because it showed how data can be manipulated for political purposes and how misleading data can have serious consequences for public health. The lack of transparency and accuracy in the reporting undermined public trust in the government’s handling of the pandemic and made it harder to implement effective public health measures.

The controversy highlighted the importance of transparency and integrity in data reporting, particularly during a public health crisis, and the need for independent verification and oversight to ensure the accuracy and reliability of data. It also underscored the responsibility of journalists and data analysts to be critical of the data they work with and to verify its accuracy and validity before reporting on it.

By avoiding these pitfalls, you can create a data story that is both compelling and trustworthy.

Common pitfalls

To create a great data story, it’s important to avoid a few common pitfalls:

1. Overcomplicating the story: While it’s important to use data and statistics to support your narrative, it’s easy to get bogged down in details and lose your audience’s attention. Try to distill your message down to its essence and communicate it clearly.
2. Cherry-picking data: It’s tempting to use only the data that supports your argument, but this can lead to a biased or incomplete story. Make sure to include all relevant data, even if it contradicts your initial hypothesis.
3. Failing to consider alternative explanations: When analyzing data, it’s important to consider all possible explanations for the patterns you observe. Don’t jump to conclusions without fully exploring alternative hypotheses.
4. Ignoring the limitations of your data: Every dataset has limitations, whether it’s due to incomplete data or sampling bias. Acknowledge these limitations and consider how they may affect your conclusions.
5. Using misleading visualizations: Visualizations are a powerful tool for conveying complex data, but they can also be misleading if not used appropriately. Make sure your visualizations accurately reflect the data and don’t distort the message you’re trying to convey.

Relevance

A great data story needs to be relevant because it needs to address a real-world problem or issue that people care about. Without relevance, the story may fail to engage the audience and may not inspire action or change. To be relevant, a data story should be based on current and important issues, use data that is recent and reliable, and provide insights or solutions that are meaningful and actionable. Additionally, it should be tailored to the audience’s interests and needs, so that they can understand the importance of the issue and relate to the story on a personal level. Overall, relevance is critical for a great data story because it provides context and meaning to the data, making it more impactful and meaningful to the audience.

Here are some examples of what can make a data story relevant:

1. Timeliness: Data stories that address current events or trends are often more relevant to audiences.
2. Localized data: Data that pertains to a specific location can be more relevant and meaningful to people who live or work in that area.
3. Personalization: Data stories that are personalized to an individual’s interests or needs can be more relevant and engaging.
4. Real-world impact: Data stories that demonstrate real-world impact or consequences can be more relevant and persuasive.
5. Novelty: Data stories that uncover new insights or challenge conventional wisdom can be more relevant and thought-provoking.
6. Connection to audience values: Data stories that align with audience values or beliefs can be more relevant and resonate more deeply.

A balanced picture

A great data story is built from many sources because it requires a diverse range of information and perspectives to fully understand the subject matter. By incorporating multiple sources, a data story can provide a more comprehensive and nuanced view of a topic, rather than relying on a single source or point of view.

For example, if a data story is about a particular social issue, it may include data from official government sources, academic research, interviews with individuals impacted by the issue, and personal narratives or testimonials. Each source contributes a different type of information and helps to create a more well-rounded and accurate story.

Furthermore, using multiple sources helps to minimize bias and ensure that the data story is not based on a single source with a particular agenda or perspective. By incorporating multiple sources, a data story can provide a more balanced and objective view of the subject matter.

How to minimize bias?

There are several ways to minimize bias in data stories:

1. Use multiple sources: As mentioned earlier, using multiple sources can help to minimize bias by providing a more balanced and objective view of the subject matter.
2. Fact-checking: Fact-checking is an important step in the data story process, as it helps to ensure that the information presented is accurate and reliable.
3. Acknowledge potential bias: If there is potential for bias in the data or the sources used, it is important to acknowledge this in the data story. This can help to provide context for the information presented and avoid misleading the audience.
4. Seek diverse perspectives: Seeking out diverse perspectives can help to ensure that the data story is not limited by a single viewpoint. This can involve interviewing a range of experts or individuals with different backgrounds and experiences.
5. Use statistical methods: Using statistical methods can help to minimize bias by providing a more objective way of analyzing the data. This can help to identify patterns or trends that may not be immediately apparent and provide a more accurate representation of the data.

Overall, minimizing bias in data stories requires a rigorous and critical approach to data analysis and presentation, as well as a commitment to transparency and objectivity.