登录查看更多内容

How can you avoid cognitive and perceptual biases when interpreting visualizations?

由人工智能和领英社区提供技术支持

Data visualization is a powerful tool to communicate complex and nuanced information, but it also comes with some challenges and pitfalls. One of them is the risk of cognitive and perceptual biases that can affect how you interpret and evaluate visualizations. These biases are mental shortcuts or distortions that can influence your judgment, memory, and reasoning. In this article, you will learn about some common types of biases and how you can avoid them when interpreting visualizations.

此文章中的业界达人

由社区从 9 条内容中精选。了解更多

Ismael Chang Ghalimi

CEO @ STOIC
Prabhát Oliveira

Data Scientist | Manager | Tech Lead @ IBM | Computer Vision | NLP | Generative AI | LLM
Michelle V.

Software Developer at AXIS Capital

1 What are cognitive and perceptual biases?

Cognitive and perceptual biases are errors or deviations from rationality that can affect how you process and understand information. They can be influenced by your personal beliefs, preferences, expectations, emotions, motivations, and social contexts. Some examples of cognitive and perceptual biases are confirmation bias, anchoring effect, framing effect, availability heuristic, and recency bias. These biases can affect how you select, analyze, and interpret data and visualizations, leading to inaccurate or misleading conclusions.

添加您的观点

Prabhát Oliveira

Data Scientist | Manager | Tech Lead @ IBM | Computer Vision | NLP | Generative AI | LLM
举报内容
By nature, we tend to analyze and interpret the information we consume with our perception of the world. In the same way, when generating content for other people to consume, our beliefs and values are loaded into the information we pass on. One of the ways to mitigate this is the contribution of different people, building their content with different opinions, in a multicultural way.

已翻译

赞
Amir Khalil Sanjani

Business Analyst- Data Analytics Professional - Financial Analyst
(已编辑)
举报内容
Remaining open to new information: is crucial in countering biases. It means being ready to change your mind when you come across fresh insights. The best way to avoid letting biases sneak in is by being willing to look at all sides of a situation.

已翻译

赞
Seetharamareddy Gowda

Data Analytics & Insights - Lead at FedEx |SQL,MS Power BI, Azure Databricks, Teradata | Mentor | Ex- Accenture | Global Sales Operations | Post Sale - Strategy & Analytics | 150 Bookings(Top 1%) on Topmate.io
举报内容
Most of the times, we have instinctive inclination towards what makes us feel good and we only look at data points which aligns with our data points. Its always good to relay on facts and insights which substantiates our views which leads to biased decision making.

已翻译

赞

2 How can you identify biases in visualizations?

To identify biases in visualizations, it is important to ask yourself some critical questions about the source, purpose, design, and message of the visualization. Who created the visualization and what is their agenda or perspective? What data was used and how was it collected, cleaned, and transformed? How was the data mapped to visual elements and what choices were made regarding the type, scale, color, shape, and layout of the visualization? What story or argument is the visualization trying to convey and what evidence or logic supports it? How does the visualization appeal to your emotions, values, or beliefs? By responding to these questions thoughtfully, you can assess the credibility, validity, and relevance of the visualization as well as any potential biases or gaps in the data or design.

添加您的观点

Michelle V.

Software Developer at AXIS Capital
举报内容
It's also important to ensure that diagrams do not just appeal to the eye but one can understand the visualization so that the viewer doesn't also have biases towards it as well. If there are any gaps in the data as it is not always the case that data quality is consistent that should be clearly illustrated to clarify why the visualizations are the way they are.

已翻译

赞
Amir Khalil Sanjani

Business Analyst- Data Analytics Professional - Financial Analyst
举报内容
Do they put more emphasis on particular aspects of the data compared to others? The creator of visualization must choose how to represent the data. These choices can include the type of visualization, the data that is included, and the colors and shapes that are used. These choices can all affect how the viewer perceives the data and can lead to the emphasis on certain aspects of the data over others. A bar chart that uses different colors for different categories of data can emphasize the differences between those categories. A pie chart that shows the relative sizes of different parts of a whole can emphasize the size of those parts. And a line graph that uses different line styles for different data series can emphasize the trends.

已翻译

赞
Seetharamareddy Gowda

Data Analytics & Insights - Lead at FedEx |SQL,MS Power BI, Azure Databricks, Teradata | Mentor | Ex- Accenture | Global Sales Operations | Post Sale - Strategy & Analytics | 150 Bookings(Top 1%) on Topmate.io
举报内容
We have to prioritize data quality in most of the data analytics projects. We need to careful on how we handle null values, spell mistakes and number formats which contributes to lot of issues in data visualization. Its always good to come up with data quality indicator dashboards for the data sources which helps to identify and fix the issue.

已翻译

赞

3 How can you avoid confirmation bias?

Confirmation bias is the tendency to seek, interpret, and remember information that confirms existing beliefs or hypotheses, while disregarding contradictory information. This can lead to overconfidence, selective attention, and confirmation traps. To avoid these traps, it is important to seek out alternative or opposing views and sources of information that challenge your assumptions and hypotheses. Additionally, it is wise to test your hypotheses with different data sets, methods, and scenarios, and look for disconfirming evidence or counterexamples. Moreover, be aware of your own biases and preferences, and strive to be objective and open-minded when interpreting visualizations. Finally, seek feedback from others who have different perspectives or expertise, and listen to their critiques and suggestions.

添加您的观点

Amir Khalil Sanjani

Business Analyst- Data Analytics Professional - Financial Analyst
举报内容
Search for disconfirming evidence or counterexamples, which can be thought of as actively seeking instances that challenge or oppose your existing notions. Additionally, by deliberately considering such disconfirming evidence or counterexamples, you can make more informed and rational decisions, nurturing an open-minded and critical thinking process.

已翻译

赞

4 How can you avoid anchoring effect?

The anchoring effect is the tendency to depend too heavily on the first piece of information encountered when making judgments or decisions, often resulting in underestimation, overestimation, or distortion of the true value or significance of the data. To prevent this, it's important to expose yourself to a variety of sources and compare and contrast different visualizations. Additionally, consider multiple explanations or outcomes for the data or situation and their probabilities and implications. Avoid making hasty decisions based on limited information and be mindful of how the presentation of information can influence your perception and evaluation. If needed, seek out more evidence or clarification.

添加您的观点

5 How can you avoid framing effect?

Framing effect is the tendency to be influenced by the way information is presented or worded, rather than by the content or meaning of the information. This can lead to different interpretations or preferences depending on how the information is framed in terms of gains or losses, risks or benefits, positive or negative, or other dimensions. To avoid framing effect, you should focus on the facts and data behind the visualization, rather than on the rhetoric or narrative that accompanies it. Rephrasing or reframing the information in different ways can help you see how it affects your understanding or attitude toward the visualization. Additionally, you should consider the context and purpose of the visualization and how it relates to your goals or interests. Finally, be mindful of how your emotions, values, or beliefs can affect your response to different frames or messages.

添加您的观点

Louis Brown

Head of Data Science | Applied AI, Data Models
举报内容
Framing is a risk that can occur when binning data for histograms or data summarisation; we need to remember that by creating the data bins, we are creating binning bias. Using deciles to split data can be an example of this. Binning is a great way of summarising data, but we need to remember that by doing this, we might be creating framing bias and should ensure that we use other metrics when working with binned data.

已翻译

赞

6 How can you avoid availability heuristic?

Availability heuristic is the tendency to judge the frequency or likelihood of an event based on how easily you can recall or imagine examples of it, rather than on the actual evidence or statistics. This can lead to overestimation, underestimation, or misperception of the prevalence or importance of the event. To avoid availability heuristic, you should seek out reliable and representative data sources and visualizations that reflect reality and diversity. Anecdotes, stories, and media reports should be avoided as they may be biased, sensationalized, or outdated. Numerical or graphical methods can be used to calculate and visualize the probability or magnitude of the event or phenomenon, and compare it with other relevant events or phenomena. Lastly, it is important to be aware of how your memory, imagination, or exposure can affect how you recall or imagine examples of the event or phenomenon.

添加您的观点

7 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?

添加您的观点

Ismael Chang Ghalimi

CEO @ STOIC
举报内容
You can avoid many cognitive and perceptual biases by simply automating your data analysis and visualization. If your data analysis and visualization tool is implemented properly, it is not subjected to the cognitive and perceptual biases that humans are subjected to, and can provide much more objective results. This is probably the most critical benefit of automation and AI.

已翻译

赞

Data Visualization

+ 关注

给文章评分

我们借助人工智能创建了此文章。您认为这篇文章怎么样？

很棒不太好

举报此文章

查看全部

How can you avoid cognitive and perceptual biases when interpreting visualizations?

1

2

3

4

5

6

7

1 What are cognitive and perceptual biases?

2 How can you identify biases in visualizations?

3 How can you avoid confirmation bias?

4 How can you avoid anchoring effect?

5 How can you avoid framing effect?

6 How can you avoid availability heuristic?

7 Here’s what else to consider

Data Visualization

给文章评分

感谢您的反馈

更多Data Visualization相关文章

更多相关阅读内容

How can you avoid cognitive and perceptual biases when interpreting visualizations?

1

2

3

4

5

6

7

1 What are cognitive and perceptual biases?

2 How can you identify biases in visualizations?

3 How can you avoid confirmation bias?

4 How can you avoid anchoring effect?

5 How can you avoid framing effect?

6 How can you avoid availability heuristic?

7 Here’s what else to consider

Data Visualization

给文章评分

感谢您的反馈

查看其他技能