(Risk 1) Risk and Uncertainty in Science Communication

Introduction to Risk and Uncertainty in Science Communication

• Discussion of risk emphasizes the inherent uncertainties in various phenomena.

• Science communication often grapples with conveying these uncertainties.

• A significant challenge arises from the misunderstanding of statistics and probability among the general public and sometimes even scientists.

The Challenge of Communicating Risk

• Effective communication about risk requires careful consideration of how probabilities and statistics are interpreted.

• Examples will clarify the complexities of conveying risk.

Example 1: Weather Forecast Probability

• A screenshot from MetService showing a weather forecast with a percentage chance of rain.

• Specifics discussed include:

  • The percentage chance of rain at 1 PM stated as 45%.

  • Questions posed to help understand this probability:

  • What does a 45% chance of rain mean?

    • Does it imply a 45% chance of being rained on at 1 PM while standing in Hamilton?

    • Does it account for the time interval of 11 AM to 3 PM (the forecast window)?

    • Does it suggest that if 100 people are surveyed at 1 PM, approximately 45 would report rain?

  • These interpretations highlight the variability and confusion surrounding how to understand probability.

  • Two scenarios with the same probability can have drastically different implications:

  • A day of scattered showers versus a large weather front affecting a region.

Example 2: Hypothetical Situation with Taxis

• A thought experiment concerning the identification of taxis based on color:

  • In a city, only taxis are pink, making up 5% of the total cars.

  • A witness claims a pink car was involved in a hit-and-run and accurately recollects car colors 98% of the time.

  • Key question raised:

  • What is the probability that the pink car was a taxi?

    • Responses to consider:

    • Is it 5% (the base rate of taxis)?

    • Is it 98% (accurate color identification)?

    • This illustrates the complexities of interpreting probabilities in real-world situations.

Example 3: Assessing Health Risks

• A scenario involving personal health:

  • A person wakes up feeling unwell and questions the probability of having COVID-19.

  • Reinforces the need for scientists to communicate risk with precision and care.

Experimental Design and Statistical Significance

• Many biologists conduct experiments comparing test and control groups:

  • Description of a trial with a drug (Wunderil) and a placebo:

  • 5,000 participants receiving the drug versus 5,000 receiving a placebo.

  • Results:

    • 730 deaths in the placebo group.

    • 664 deaths in the Wunderil group.

  • Preliminary analysis indicates a statistically significant p-value of 0.004.

  • Discussion points:

  • While statistically significant, the percentage decrease in death rates is minimal:

    • Death rate changes from approximately 14% (placebo) to 13.25% (Wunderil).

  • Reinforces the distinction between statistical significance and practical significance.

Importance of Practical Significance

• Highlight that statistical significance does not automatically translate into practical relevance:

  • Significant medical treatments must also justify their costs versus real-world benefits.

  • Miscommunication often arises when scientists focus solely on statistical outcomes without addressing practical implications.

Summary and Transition to Next Video

• The introduction outlined concerns about effectively communicating risk and uncertainty.

• The next part of the series will explore structures for discussing risk and uncertainty meaningfully, building on the challenges identified.