Probability Distributions - Summary Notes

Lesson Overview

• Introduction to Probability Distributions: A probability distribution is a mathematical function that models the likelihood of all possible outcomes for a random variable. It provides a structured way to analyze data variability and predict future events.

• Classification of Distributions:

  • Discrete Probability Distributions: These represent data that can be counted (integers). Examples include:

  • Binomial Distribution: Models the number of successes in $n$ independent trials (e.g., coin flips).

  • Poisson Distribution: Models the number of events in a specified interval (e.g., number of emails per hour).

  • Continuous Probability Distributions: These represent data that can take any value within a range (measured). Examples include:

  • Normal Distribution: Often called the 'Gaussian distribution,' it is characterized by its bell-shaped curve.

  • Exponential Distribution: Describes the time between independent events.

• Key Statistical Parameters:

  • Mean ($\mu$): The arithmetic average, representing the distribution's center.

  • Variance ($\sigma^2$): The average squared deviation from the mean, quantifying total dispersion.

  • Standard Deviation ($\sigma$): The square root of variance, mapping the spread back to the original unit scale.

Characteristics of the Normal Distribution

• Visual Shape: It is a unimodal, bell-shaped curve that is perfectly symmetric around its mean ($\mu$).

• Central Tendency: In a perfectly normal distribution, the Mean, Median, and Mode are all equal and located at the center peak.

• Asymptotic Nature: The curve's tails approach the horizontal axis as they move away from the mean but never touch it, extending from $-\infty$ to $+\infty$.

• Total area Under the Curve: The total area is exactly $1$ (or 100%), representing the total probability of all possible outcomes.

The Empirical Rule (68-95-99.7)

In a normal distribution, the following probabilities apply based on the distance from the mean:

• Approx. 68.27% of data points lie within one standard deviation: $P(\mu - \sigma \leq X \leq \mu + \sigma)$.

• Approx. 95.45% of data points lie within two standard deviations: $P(\mu - 2\sigma \leq X \leq \mu + 2\sigma)$.

• Approx. 99.73% of data points lie within three standard deviations: $P(\mu - 3\sigma \leq X \leq \mu + 3\sigma)$.

Standard Normal Distribution and Z-Scores

• The Concept of Standardization: To analyze different normal distributions on a unified scale, they are transformed into a Standard Normal Distribution where $\mu = 0$ and $\sigma = 1$.

• The Z-Score Formula: This value represents how many standard deviations an observation ($X$) is from the mean. It is calculated as:
  $Z = \frac{X - \mu}{\sigma}$

• Utility: Standardizing allows researchers to use a standard normal table ($Z$-table) to find the cumulative probability (the area under the curve) to the left of any given point.

The Central Limit Theorem (CLT)

• Definition: The CLT states that the sampling distribution of the sample mean will approach a normal distribution as the sample size increases, regardless of the original population distribution shape.

• Significance: This theorem is the foundation for much of inferential statistics, allowing the use of normal distribution methods for variables that may not be normally distributed individually.

Real-World Applications and Examples

• Biological Measurements: Human heights, blood pressure, and birth weights often follow a normal distribution, where most individuals cluster around the mean.

• Psychology and Education: IQ scores are designed with a mean of $100$ and a standard deviation of $15$ to ensure a predictable distribution of scores across the population.

• Errors in Measurement: Random errors in scientific experiments typically distribute themselves according to a normal curve.