Once upon a time in the country of Data Land, there were two cities called "Average City" and "Median City." Each city measured the happiness levels of its citizens, but they used different methods to do so.
In Average City, they calculated the average by adding up all citizens' happiness scores and dividing by the total number of people. This method captures an overall trend of everyone’s happiness. In contrast, Median City measured happiness by ordering citizens from the least to most happy and selecting the happiness score of the person exactly in the middle. The median is less influenced by outliers, such as extremely happy or unhappy individuals.
Thus, the two cities learned the importance of interpreting data through their distinct measurement methods. However, this is only the beginning of Data Land's story. A new challenge was about to unfold in the two cities.
One hot summer day, Average City saw a significant increase in ice cream sales, as citizens, feeling the intense heat, flocked to buy refreshing ice cream. At the same time, more people began swimming to cool off in local pools. Unfortunately, this also led to an increase in swimming accidents. Noticing this odd phenomenon, scientists in the city observed a "correlation" between ice cream sales and swimming accidents. Despite the seemingly unrelated nature of these two events, their connection was dubbed the "Ice Cream-Swimming Correlation" and quickly became a topic of interest among the citizens.
However, some citizens misunderstood this "correlation." Rumors spread that "eating ice cream increases the risk of swimming accidents," leading many to avoid ice cream. Consequently, ice cream sales plummeted, leaving shop owners perplexed. To resolve the situation, Mr. Average of Average City and Mr. Median of Median City decided to hold a public discussion to explain the difference between correlation and causation to their citizens.
During the discussion, the scientists explained, "A correlation means that two events tend to occur together, but it doesn't necessarily mean one causes the other. On the other hand, causation means that one event directly causes the other.”
The scientists elaborated further, revealing that there was no direct causal link between ice cream sales and swimming accidents. In reality, an invisible factor, the "temperature," influenced both events. On hot days, people crave more ice cream and also enjoy swimming, which unfortunately leads to an increase in accidents. This hidden variable was actually influencing both events.
Upon hearing this explanation, the citizens began to understand that eating ice cream does not directly cause swimming accidents. They resumed eating ice cream without worry, and ice cream sales quickly bounced back. Citizens could now enjoy cooling off with ice cream and swimming safely on hot days.
Through this experience, the citizens of Data Land learned that data is more than just numbers; each data point can tell its own story. They realized that just because two events happen at the same time, it doesn’t automatically mean one causes the other. They also understood the importance of considering hidden factors and other influences when interpreting data.
Explanation: The Story of Data Land No. 2: Learning About Correlation and Causation from Ice Cream - The Story Behind the Numbers
Through the story of Data Land, we can see the importance of understanding the difference between correlation and causation. In Data Land, there were two cities, “Average City” and “Median City,” each with its own way of measuring happiness. Average City used the "average" method, while Median City used the "median." These differing approaches highlight the importance of perspective in data interpretation.
In this story, Average City encountered a strange phenomenon where both ice cream sales and swimming accidents increased at the same time during hot summer days. This pattern was influenced by a hidden variable, temperature, as hot days caused people to crave ice cream and go swimming more frequently, thereby increasing the likelihood of accidents. However, some citizens mistakenly believed that ice cream consumption directly caused the increase in swimming accidents.
To clear up this misunderstanding, a public discussion was held where scientists explained the difference between correlation and causation. They clarified that just because two events occur simultaneously does not mean one causes the other. Instead, other hidden factors, such as “temperature” in this case, could influence both events. After hearing this, the citizens recognized the misunderstanding, no longer believed the rumor, and resumed eating ice cream.
Through this story, the citizens of Data Land learned that data is more than just numbers and that interpreting data accurately often requires considering hidden factors and other influences. The knowledge that correlation does not imply causation is an essential insight in data science.
The story of Data Land reminds us of the importance of recognizing "hidden factors" behind observed correlations. By understanding these principles, we can develop the skills to interpret data correctly and view everyday information from multiple angles.
