Chapter 8 – Post Hoc Rides Again
In this chapter, we explore the fallacy of assuming that correlation implies causation, often summarized by the Latin phrase “post hoc ergo propter hoc” (after this, therefore because of this). The chapter emphasizes that just because two events or phenomena occur together doesn’t mean one caused the other.
Key Points:
- Correlation vs. Causation: The chapter stresses the crucial distinction between correlation (two things happening together) and causation (one thing directly causing another). It explains that while correlation can be a starting point for investigation, it doesn’t automatically prove a cause-and-effect relationship.
- Example: A coastal town observes a correlation between sunscreen sales and shark attacks. As sunscreen sales increase, so does the number of shark attacks. A misleading conclusion could be that sunscreen attracts sharks or increases the likelihood of being attacked. However, this correlation doesn’t imply causation. The underlying factor is likely the summer season. During summer, more people buy sunscreen to protect themselves from the sun, and more people go swimming in the ocean, increasing the chances of encountering sharks.
- Confounding Factors: Often, a third, unseen factor is responsible for the apparent connection between two things. This hidden variable is called a confounding factor. It’s essential to consider and rule out potential confounding factors before concluding that one thing causes another.
- Example: A study might find that people who take vitamin supplements are healthier. However, this doesn’t necessarily mean the vitamins are directly causing better health. The confounding factor could be that people who take vitamins are also more likely to engage in other healthy behaviors, like exercising and eating a balanced diet.
- Reverse Causation: Sometimes, the assumed direction of causation might be reversed. The chapter explains that it’s possible for the supposed effect to actually be the cause.
- Example: A study might find that people who are hospitalized are more likely to die. However, this doesn’t mean hospitals are inherently dangerous. The reverse causation is that people are hospitalized because they are already sick or injured, which increases their risk of death.
- Coincidence: Sometimes, the correlation between two events is purely coincidental. There’s no causal relationship, and the events just happen to occur together by chance.
- Example: A study might find that people who wear a certain brand of shoes are more successful in their careers. However, this correlation is likely coincidental. There’s no plausible reason why a particular brand of shoes would directly impact career success.
The Importance of Critical Thinking:
The chapter emphasizes the importance of critical thinking when evaluating claims of causation. Readers should question whether a correlation truly implies causation, consider potential confounding factors, and be open to the possibility of reverse causation or coincidence. By carefully examining the evidence and considering alternative explanations, readers can avoid being misled by false claims of causation.