Conditional Probability

Prepared by : Reband Kawa Supervised by : [Link] Bayan

Lazo Rayar
Vanessa Bahman
 Introduction to Probability

• Probability measures uncertainty – It quantifies how likely an

event is to occur, from impossible (0) to certain (1).
• Values range from 0 to 1 – 0 means no chance, 1 means
absolute certainty; values in between reflect degrees of
likelihood (e.g., 0.5 = even chance).
• Used in science, medicine, economics, daily life – Examples:
testing drug efficacy, predicting market trends, weather
forecasting, or deciding whether to carry an umbrella.
• .
 Types of Probability

• Theoretical Probability – Based on known possible outcomes

without experimentation. Calculated as: favorable outcomes / total
outcomes (e.g., probability of rolling a 3 on a fair die = 1/6).
• Empirical (Frequentist) Probability – Based on observed data from
experiments or past events. Calculated as: number of times event
occurred / total trials (e.g., if it rained 30 of 100 days, empirical
probability of rain = 0.3).
• Subjective Probability – Based on personal judgment, experience, or
belief rather than exact calculation (e.g., “I think there’s a 70% chance
the stock market will rise tomorrow”)
 Why Probability Matters

• Helps make informed decisions – Probability provides a rational

basis for choosing between options by weighing the likelihood of
different outcomes (e.g., deciding whether to undergo a medical
procedure based on success rates).
• Quantifies risk and uncertainty – It assigns numerical values to
risks, allowing comparison and management (e.g., insurance
premiums calculated from accident probabilities, or investment risks
in finance).
• Supports predictions and testing reliability – Probability underpins
statistical inference, enabling predictions from samples and
assessing the reliability of results (e.g., poll predictions, quality
control in manufacturing, and p-values in scientific studies).
 Probability in Weather Forecasting

• Modern forecasting uses probabilistic methods – Instead of single-

outcome predictions, forecasters use ensemble models (running
many simulations with slight variations) to generate a range of
possible future weather scenarios.
• Atmosphere is chaotic and uncertain – Due to the butterfly effect,
small measurement errors or unknown variables can grow over time,
making exact long-term prediction impossible.
• Forecasts now use percentages instead of yes/no answers – For
example, a "70% chance of rain" means that in 7 out of 10 similar
atmospheric conditions, rain occurred—better communicating
uncertainty and helping users make risk-based decisions (e.g.,
carrying an umbrella vs. canceling an outdoor event).
 What Does 30% Chance of Rain
Mean?

• Frequentist interpretation – The "30% chance of rain" means that over

many past days with identical atmospheric conditions, it rained on only
30% of those days. Probability is defined as long-run relative frequency.
• Ensemble model interpretation – A forecast model runs multiple
simulations (typically 20–50) with slightly different initial conditions. If
rain occurs in 30% of those ensemble members, the forecast gives a 30%
probability. Probability reflects model agreement.
• Subjective interpretation by meteorologists – Forecasters combine
model outputs, local knowledge, and experience to adjust probabilities.
For example, they might lower rain chance from 50% to 30% based on
drying trends observed on radar. Probability represents a degree of belief
or confidence.
 Conditional Probability in Weather

• Rain depends on humidity, pressure, and temperature – These are

interacting meteorological variables. For rain to occur, conditions
often include high humidity, low pressure, and specific temperature
ranges.
• Example: P(Rain | Humidity > 80%) = 50% – This is a conditional
probability: given that humidity exceeds 80%, the chance of rain
becomes 50%, which is typically higher than the unconditional
(baseline) probability of rain.
• Conditional probability improves forecasts – By conditioning on
known or measured variables (e.g., current humidity), forecasters
update probabilities dynamically, making predictions more accurate
and location-/time-specific than static averages.
 Probability in Medical Testing

• Medical tests are imperfect – No diagnostic test is 100% accurate.

Errors can occur due to biological variability, test limitations, or sample
contamination.
• False positives and false negatives occur – A false positive means the
test indicates a disease when it is not present. A false negative means the
test indicates no disease when it is actually present.
• Sensitivity and specificity measure accuracy – Sensitivity is the true
positive rate (ability to correctly identify those with the
disease). Specificity is the true negative rate (ability to correctly identify
those without the disease). These help evaluate and compare test
performance.
• .
 Bayes' Theorem

• Updates probabilities using new evidence – Bayesian reasoning starts

with a prior probability (initial belief) and revises it after observing new
data to produce a posterior probability.
• Used to interpret medical test results – Even with a highly accurate
test, the probability of actually having a disease depends on both the test
result and the baseline prevalence. Bayes' theorem calculates this
properly.
• Important for rare diseases – For a rare disease, even a positive test
result can mean a low probability of actually having the disease, because
false positives may outnumber true positives. Example: if a disease
affects 1 in 10,000 people, a 99% accurate test will yield many more false
positives than true positives, so P(Disease | Positive) might be only ~1%.
 Weather vs. Medicine

• Both rely on conditional probability – Weather forecasts use P(Rain |

Humidity, Pressure) and medical tests use P(Disease | Test Result). Both
update the probability of an event given new evidence.
• Both face public misunderstanding – People often misinterpret a 30%
rain chance as "it will rain 30% of the day" rather than "30% of similar
days had rain." Similarly, patients mistake a positive test result as near-
certain disease, ignoring prevalence and false positives.
• Medicine uses Bayes' theorem more explicitly – While meteorologists
apply Bayesian thinking informally (e.g., updating forecasts with radar),
medicine explicitly teaches and calculates Bayes' theorem for screening
decisions, diagnostic reasoning, and communicating post-test
probabilities to patients.
 Conclusion

• Probability helps manage uncertainty – Instead of demanding absolute

certainty, probability provides a quantitative framework for making
rational choices when outcomes are unknown.
• Base rates are extremely important – Ignoring how common (or rare)
an event is in the population (e.g., disease prevalence, prior crime rates)
leads to serious errors, as seen in medical testing and legal judgments.
Base rates anchor all conditional probability updates.
• Probabilistic thinking leads to better decisions – Recognizing that
outcomes involve chance, not certainty, reduces overconfidence,
improves risk assessment, and helps avoid black-and-white thinking—
whether in finance, medicine, weather, or daily life.
 References

• Reference: Gigerenzer, G., Gaissmaier, W., Kurz-Milcke, E., Schwartz, L. M., &
Woloshin, S. (2007).
• Reference: Mylne, K., & Roberts, H. (2024).
• Reference: Kahneman, D., & Tversky, A. (1972). Subjective probability