Statistical Thinking in Science: Crash Course Scientific Thinking #2

Understanding Statistics in Everyday Life

Key Concepts:

• Mean: The average value, calculated by summing all values and dividing by the number of values.
• Median: The middle value in a dataset when arranged in order.
• Mode: The most frequently occurring value in a dataset.
• Standard Deviation: A measure of how spread out data points are from the mean.
• Confidence Interval: A range of values within which a true value is likely to fall, with a specified level of confidence.
• Relative Risk: The increase in risk compared to a baseline.
• Absolute Risk: The actual probability of an event occurring.
• Correlation: A statistical relationship between two or more variables.
• Causation: A relationship where one variable directly influences another.
• Confounding Variable: A factor that influences both the variables being studied, creating a spurious association.
• Statistical Significance: The likelihood that a result is not due to random chance.

Determining Typical Age of Death

The video begins by questioning how to determine a “typical” age of death for an American man. While a national dataset indicates an average (mean) age of death of 70 years between 2018-2023, this number can be misleading. The mean is susceptible to being skewed downwards by deaths occurring at younger ages.

Alternatively, the most common age of death (mode) is 79. However, the majority of deaths actually occur below 79, meaning it’s more probable an individual would die at an age lower than the mode.

To gain a more nuanced understanding, the video introduces the concept of standard deviation. This metric reveals how spread out the data points are from the mean, providing insight into how “typical” a particular age is. The median age of death (73) is also presented as a potentially more representative measure, as it’s less susceptible to skewing than the mean. The key takeaway is that different measures of central tendency (mean, median, mode) offer different perspectives, and combining them with standard deviation provides a more complete picture.

The Importance of Confidence Intervals

The video emphasizes that statistics inherently involve uncertainty. To assess the reliability of a statistic, it’s crucial to consider the confidence interval. A 95% confidence interval, for example, indicates that if the study were repeated 100 times with new samples, the statistic would fall within that range approximately 95 times. This highlights the inherent variability in statistical results and the importance of understanding the precision of a given number. Hank Green stresses that “it is way better to be roughly right than precisely wrong.”

Relative vs. Absolute Risk: The Birth Control Pill Example

A compelling case study involves a new birth control pill reported to increase the risk of blood clots by 100%. This sounds alarming, but the video clarifies the distinction between relative risk and absolute risk. The 100% increase refers to relative risk – the risk doubled from 1 in 7,000 to 2 in 7,000.

The absolute risk remains low. The video points out that pregnancy itself carries a higher risk of blood clots, demonstrating how understanding absolute risk is crucial for informed decision-making. This example underscores how statistics can be manipulated or misinterpreted to create a misleading impression.

Correlation vs. Causation & Confounding Variables

The video then explores the concept of correlation, defining it as a relationship between variables. The strength of this relationship is quantified by the R value, ranging from -1 to 1. A value of 1 indicates a perfect positive correlation, -1 a perfect negative correlation, and 0 indicates no correlation.

However, the crucial point is that correlation does not equal causation. The example of sunscreen use and skin cancer is used to illustrate this. While a strong negative correlation exists (higher sunscreen use correlates with lower skin cancer rates), there’s also evidence of a causal link – wearing sunscreen does reduce cancer risk.

Conversely, the correlation between ice cream sales and shark attacks is presented as an example of a spurious correlation driven by a confounding variable – warm weather. Warm weather leads to both increased ice cream consumption and more people swimming, increasing the likelihood of shark encounters.

The video further explains that even when controlling for confounding variables, results might be due to random chance, necessitating an assessment of statistical significance.

Statistical Significance vs. Real-World Importance

Finally, the video clarifies that statistical significance doesn’t necessarily equate to real-world importance. A statistically significant result simply means it’s unlikely to have occurred by chance, but it doesn’t guarantee the finding is meaningful or impactful. Hank Green humorously illustrates this with the example of Doritos being a “significant” part of his life, highlighting the difference between statistical and personal significance.

Conclusion:

The video emphasizes the importance of critical thinking when encountering statistics. Understanding concepts like mean, median, mode, standard deviation, confidence intervals, relative vs. absolute risk, correlation vs. causation, and statistical significance empowers individuals to interpret data accurately and make informed decisions. The core message is that statistics are powerful tools, but their value lies in understanding their limitations and interpreting them within the appropriate context. Acknowledging inherent uncertainty and digging deeper beyond surface-level numbers are essential for navigating a world increasingly reliant on data.