Correlation is Not Causation

“Correlation is not causation” is a critical concept in statistics and research methodology, reminding us that just because two variables are related does not necessarily mean that one causes the other.

The nature of causality is systematically investigated in several academic disciplines, including philosophy and physics. It is drilled, military school-style, into every budding statistician. Just because two things correlate does not necessarily mean that one causes the other. As a seasonal example, just because people in the UK tend to spend more in the shops when it’s cold and less when it’s hot doesn’t mean cold weather causes frenzied high-street spending. A more plausible explanation would be that cold weather tends to coincide with Christmas and New Year sales.

Some Examples of “Correlation is Not Causation”

  • Ice cream sales and drowning incidents: During hot summer months, ice cream sales tend to increase, and at the same time, the number of drowning incidents rises. This does not mean that eating ice cream causes people to drown or vice versa. The correlation between the two variables is due to a confounding factor, i.e., hot weather, which increases both ice cream sales and people swimming, leading to more drownings.
  • Number of firefighters and fire-related damages: In a city with a larger number of firefighters, there is often a higher incidence of fire-related damages. However, hiring more firefighters does not cause an increase in fires; rather, cities with higher fire risks tend to employ more firefighters to cope with potential emergencies.
  • Education level and income: Studies consistently show that higher education levels are correlated with higher incomes. Still, it would be misleading to assume that getting a college degree directly causes someone to earn more money. Other factors, such as personal skills, work experience, and career opportunities, also play a significant role in determining income levels.
  • Coffee consumption and longevity: Some studies suggest that moderate coffee consumption is correlated with increased lifespan. However, this does not mean that drinking coffee alone makes people live longer. Other factors, such as overall lifestyle, diet, and genetics, are more likely to influence life expectancy.
  • TV watching and obesity: Research has found a positive correlation between the amount of time spent watching television and obesity rates. Yet, this does not mean that watching TV directly causes people to become obese. Sedentary behavior and unhealthy eating habits associated with excessive TV watching are more likely contributing factors to obesity.
  • Alcohol consumption and intelligence: Some studies have shown a negative correlation between alcohol consumption and intelligence. However, this does not mean that drinking alcohol lowers intelligence; rather, other factors such as socioeconomic status or lifestyle choices could be involved.
  • Birth control usage and poverty rates: Countries with higher rates of birth control usage might also have lower poverty rates. However, this does not imply that using birth control causes a decrease in poverty; instead, it could be that countries with better economic conditions tend to adopt family planning methods.
  • Number of storks and birth rates: In some regions, there is a humorous correlation between the number of storks observed and birth rates. However, storks do not deliver babies, and this correlation is due to common factors such as rural environments with larger families and more stork habitats.
  • Amount of sleep and academic performance: Research has shown a positive correlation between getting more sleep and better academic performance. However, it is not accurate to conclude that increasing sleep duration directly improves grades; there may be other variables, such as study habits or motivation, influencing academic success.
  • Car color and accident rates: Some studies have found a correlation between certain car colors and accident rates. However, car color is not a causative factor for accidents; instead, the correlation might be due to other factors like driving behavior, road conditions, and the number of cars of a specific color on the road.
  • Internet usage and depression: Some studies suggest a positive correlation between internet usage and depression. However, this does not mean that using the internet causes depression; rather, individuals experiencing depression might be more inclined to spend more time online.
  • Child height and reading ability: Research has shown a positive correlation between a child’s height and reading ability. However, taller children are not inherently better readers; this correlation might be influenced by age-related growth patterns and the fact that older children are generally taller and more proficient readers.
  • Coffee consumption and stress: Some studies indicate that coffee consumption is negatively correlated with stress levels. However, this does not imply that drinking more coffee reduces stress; rather, it could be that individuals with lower stress levels feel more comfortable consuming coffee.
  • Internet speed and economic growth: Countries with faster internet speeds might experience higher economic growth rates. However, fast internet does not directly cause economic growth; rather, countries with better economies can afford to invest in advanced internet infrastructure.
  • Chocolate consumption and Nobel laureates: Some studies have humorously correlated chocolate consumption per capita with the number of Nobel laureates in a country. However, eating more chocolate does not lead to producing more Nobel laureates; this correlation is coincidental and does not imply causation.
  • Sunscreen usage and sunburns: People who use sunscreen regularly might still experience sunburns. While sunscreen helps protect against UV rays, its effectiveness depends on factors like proper application, sun exposure duration, and individual skin sensitivity.
  • Social media usage and self-esteem: Studies have found a negative correlation between social media usage and self-esteem. However, social media usage is not a direct cause of low self-esteem; rather, individuals with lower self-esteem might be more likely to seek validation and engagement on social platforms.
  • Number of books at home and academic success: Research has shown a positive correlation between the number of books at home and a child’s academic success. However, simply having more books does not cause better academic performance; instead, a home with more books might indicate a more intellectually stimulating environment.
  • Number of Facebook friends and real-life friendships: While there is a correlation between the number of Facebook friends and the size of one’s real-life social network, it does not indicate causation. Social media connections may not necessarily reflect the depth of offline friendships.
  • Number of lawyers and lawsuit frequency: A higher number of lawyers in a region does not directly lead to more lawsuits. The correlation might be influenced by population size, legal complexity, and other legal factors.
  • Number of police officers and crime rates: The number of police officers in an area does not solely determine crime rates. Crime is influenced by various factors, such as socioeconomic conditions, community engagement, and law enforcement strategies.
  • Distance to the equator and skin cancer rates: While there is a correlation between distance from the equator and skin cancer rates, it’s not the distance alone that causes higher skin cancer rates. Exposure to UV radiation and sun protection habits are crucial factors.
  • Average shoe size and mathematical ability: A positive correlation between shoe size and mathematical ability does not mean that having larger feet enhances math skills. The correlation might be coincidental and unrelated.
  • GDP per capita and chocolate consumption: Countries with higher GDP per capita do not necessarily have higher chocolate consumption due to wealth alone. The correlation could be influenced by cultural preferences, dietary habits, and availability.
  • Number of supermarkets and obesity rates: A higher number of supermarkets in an area does not necessarily cause higher obesity rates. The correlation might be due to supermarkets locating in areas with larger populations, which could include individuals with different dietary habits.
  • Exercise frequency and weight loss: While regular exercise is associated with weight loss, correlation does not mean causation. Weight loss is a complex process influenced by factors like diet, metabolism, and genetics.
  • Religious attendance and morality: Correlations between religious attendance and certain moral behaviors do not imply causation. Other variables, such as cultural norms, upbringing, and individual beliefs, also influence moral behavior.
  • Number of fire engines and fire damage: The presence of more fire engines in a city does not cause an increase in fire damage. Fire departments station more vehicles in areas with higher risks to mitigate potential losses.
  • Social media usage and happiness: While some studies have found a negative correlation between heavy social media usage and happiness, it’s important to note that causation cannot be established. Other factors, such as personal circumstances and offline relationships, play a significant role in overall happiness.
  • Employment rate and political party in power: Changes in the employment rate during a political party’s term do not necessarily prove that their policies directly caused the fluctuations. Economic trends are influenced by various factors beyond the control of a single political party.
  • Number of physicians and life expectancy: Countries with more physicians do not necessarily have longer life expectancies solely due to the number of doctors. Life expectancy is influenced by a complex interplay of factors, including healthcare infrastructure, lifestyle choices, and social determinants of health.
  • Smartphone ownership and intelligence: Higher smartphone ownership rates in a country do not directly lead to higher intelligence levels among its population.

Reverse causation, also known as reverse causality or wrong direction, is a logical fallacy that occurs when the cause-and-effect relationship between two variables is mistakenly reversed. It’s essential to be cautious when interpreting correlations and avoid falling into the trap of reverse causation, as it can lead to misleading conclusions about cause-and-effect relationships.

Some Examples of Reverse Causation
  • Crime and police presence: Claiming that a high police presence causes crime, instead of understanding that crime rates may lead to increased police presence in an area.
  • Healthy lifestyle and exercise: Arguing that engaging in physical activities leads to a healthy lifestyle, while, in reality, individuals with healthier lifestyles are more likely to participate in regular exercise.
  • Economic growth and government spending: Suggesting that increased government spending causes economic growth, while it could be that economic growth allows for higher government spending due to increased tax revenues.
  • Education level and intelligence: Wrongly assuming that higher intelligence results from obtaining higher levels of education, when in fact, individuals with higher intelligence may be more likely to pursue advanced education.
  • Internet usage and social skills: Believing that excessive internet usage leads to poorer social skills, it could be that individuals with weaker social skills are more drawn to spending time online.
  • The faster that windmills are observed to rotate, the more wind is observed. Therefore, wind is caused by the rotation of windmills. (Or, simply put windmills, as their name indicates, are machines used to produce wind.)
  • In other cases, it may simply be unclear which is the cause and which is the effect. For example, children that watch a lot of TV are the most violent. Clearly, TV makes children more violent. This could easily be the other way around; that is, violent children like watching more TV than less violent ones.

The third-cause fallacy, also known as ignoring a common cause or questionable cause, occurs when a causal relationship is mistakenly attributed to two variables (X and Y) when, in reality, both X and Y are caused by a third variable (Z).

This indicates the importance of considering alternative explanations and avoiding hasty conclusions about causation when observing correlations between variables. It is essential to thoroughly investigate the underlying mechanisms and potential confounding factors to avoid falling into the third-cause fallacy.

Examples of Ignoring a Common Cause
  • Ice cream sales and drowning incidents: Assuming that the increase in ice cream sales causes an increase in drowning incidents, when, in fact, both are influenced by a common cause, such as hot weather during summer months.
  • Education level and income: Mistakenly concluding that higher education levels cause higher income, without considering that both education and income are influenced by factors like intelligence, skills, and job opportunities.
  • Coffee consumption and stress: Believing that drinking more coffee causes higher stress levels, without acknowledging that both coffee consumption and stress might be responses to a busy and stressful lifestyle.
  • Number of firefighters and fire-related damages: Incorrectly assuming that an increase in the number of firefighters leads to more fire-related damages, without considering that areas with higher fire risks may have more firefighters to address potential emergencies.
  • Vaccination and autism: Erroneously associating vaccines with autism, when the correlation arises due to a common cause, such as genetic factors, that may influence both autism rates and vaccination decisions.
  • Sleeping with one’s shoes on is strongly correlated with waking up with a headache. Therefore, sleeping with one’s shoes on causes headaches.
  • Young children who sleep with the light on are much more likely to develop myopia in later life. Therefore, sleeping with the light on causes myopia.
  • Since the 1950s, both the atmospheric CO2 level and obesity levels have increased sharply. Hence, atmospheric CO2 causes obesity.
Further Reading
Sources

Author: Doyle

I was born in Atlanta, moved to Alpharetta at 4, lived there for 53 years and moved to Decatur in 2016. I've worked at such places as Richway, North Fulton Medical Center, Management Science America (Computer Tech/Project Manager) and Stacy's Compounding Pharmacy (Pharmacy Tech).

Leave a Reply

Discover more from Doyle's Space

Subscribe now to keep reading and get access to the full archive.

Continue reading