Mehmet Sat | Personal Blog

Probability Distributions

· mehmet sat

Let’s dive into the world of probability distributions. Imagine you’re trying to understand randomness. Probability distributions are like the blueprints or recipes that describe how randomness behaves in different situations. They are a fundamental concept in probability and statistics and are used everywhere from weather forecasting to financial modeling to understanding the spread of diseases.

What is a Probability Distribution? (The Core Idea)

At its heart, a probability distribution is a complete description of the probabilities of all possible outcomes for a random variable.

Let’s break this down:

  • Random Variable: This is a variable whose value is a numerical outcome of a random phenomenon. Think of it as something you can measure or count that is subject to chance.

    • Examples of Random Variables:
      • The number you get when you roll a die.
      • The height of a randomly selected person.
      • Whether it will rain tomorrow (we can assign 1 for rain, 0 for no rain).
      • The time it takes for a light bulb to burn out.

  • Possible Outcomes: These are all the potential values the random variable can take.

    • Examples of Possible Outcomes:
      • For a die roll: 1, 2, 3, 4, 5, 6.
      • For height: Any value within a reasonable range (e.g., 0 to 3 meters, though practically more limited).
      • For rain tomorrow: Rain (1) or No rain (0).
      • For light bulb lifespan: Any positive time value (theoretically).

  • Probabilities: For each possible outcome, a probability distribution tells you the likelihood of that outcome occurring. Probabilities are always numbers between 0 and 1 (or 0% to 100%).

Think of it like this:

Imagine you have a bag of marbles. If you know the exact number of marbles of each color in the bag, you have a “distribution” of colors. A probability distribution is similar, but instead of colors, we’re dealing with the possible values of a random variable, and instead of counts, we’re dealing with probabilities.

Two Main Types of Probability Distributions

Probability distributions come in two main flavors, depending on the type of random variable they describe:

  1. Discrete Probability Distributions: These deal with random variables that can only take on a countable number of values. Think of things you can count, usually whole numbers.

    • Key Characteristic: You can list out all the possible values and assign a probability to each. The probabilities must all add up to 1 (or 100%).

    • Examples of Discrete Distributions:

      • Bernoulli Distribution: (Think: Single Coin Flip)

        • Random Variable: X = Result of a single coin flip (e.g., 0 = Tails, 1 = Heads)
        • Possible Outcomes: 0, 1
        • Probability Distribution:
          • P(X = 0) = Probability of Tails (let’s say ‘p’ for heads, so 1-p for tails)
          • P(X = 1) = Probability of Heads (p)
        • Example: For a fair coin, p = 0.5. So, P(X=0) = 0.5 and P(X=1) = 0.5.

      • Binomial Distribution: (Think: Multiple Coin Flips or Trials)

        • Random Variable: X = Number of successes in a fixed number of independent trials.
        • Possible Outcomes: 0, 1, 2, …, n (where ’n’ is the number of trials)
        • Probability Distribution: Formula to calculate the probability of getting exactly ‘k’ successes in ’n’ trials, given the probability of success on a single trial (p).
        • Example: Flipping a coin 3 times. What’s the probability of getting exactly 2 heads? The binomial distribution can tell you this.

      • Poisson Distribution: (Think: Rare Events over Time or Space)

        • Random Variable: X = Number of events occurring in a fixed interval of time or space.
        • Possible Outcomes: 0, 1, 2, 3, … (theoretically infinite, but probabilities get very small for large numbers)
        • Probability Distribution: Formula based on the average rate of events (λ - lambda).
        • Example: Number of customers arriving at a store per hour, number of typos on a page, number of accidents at an intersection per year.

      • Discrete Uniform Distribution: (Think: Fair Die)

        • Random Variable: X = Outcome of rolling a fair die.
        • Possible Outcomes: 1, 2, 3, 4, 5, 6
        • Probability Distribution: Each outcome has an equal probability. For a fair 6-sided die, P(X=1) = P(X=2) = … = P(X=6) = 1/6.

  2. Continuous Probability Distributions: These deal with random variables that can take on any value within a given range (or even the entire real number line). Think of things you measure rather than count.

    • Key Characteristic: Instead of probabilities for specific values, we talk about probabilities of the random variable falling within a certain range of values. We use a probability density function (PDF) to describe the distribution. The area under the PDF curve within a range represents the probability of the variable falling in that range.

    • Examples of Continuous Distributions:

      • Normal Distribution (Gaussian Distribution): (Think: Bell Curve - Very Common in Nature)

        • Random Variable: Many things in nature and human measurements tend to follow a normal distribution (approximately).
        • Shape: Symmetrical bell shape, centered around the mean. Defined by its mean (μ - mu) and standard deviation (σ - sigma).
        • Examples: Height of adult males, blood pressure, exam scores (often, though not always perfectly).

      • Uniform Distribution (Continuous): (Think: Random Number Generator within a Range)

        • Random Variable: Any value within a specified range is equally likely.
        • Shape: Flat, rectangular shape.
        • Example: The time a bus arrives at a stop if it’s supposed to arrive uniformly between 9:00 AM and 9:15 AM.

      • Exponential Distribution: (Think: Time Until an Event - Often for Waiting Times)

        • Random Variable: Time until an event occurs.
        • Shape: Decreasing curve, highest probability at zero and decreasing as time increases.
        • Examples: Lifespan of electronic components, time between customer arrivals at a service counter, time until a radioactive atom decays.

Visualizing Probability Distributions

  • Discrete Distributions: Often visualized with bar charts or histograms. The height of each bar represents the probability of that specific outcome.

  • Continuous Distributions: Visualized with curves. The area under the curve between two points represents the probability of the random variable falling within that range.

Why are Probability Distributions Important?

  • Understanding Randomness: They give us a framework to understand and model random phenomena.
  • Making Predictions: We can use them to predict the likelihood of future events. For example, weather forecasting uses probability distributions to predict the chance of rain.
  • Decision Making: In business, finance, and many other fields, probability distributions help in making informed decisions under uncertainty. For example, assessing the risk of an investment.
  • Statistical Inference: They are the foundation for statistical tests and analysis. We use them to draw conclusions about populations based on sample data.

Example in Action: Normal Distribution of Heights

Let’s say we know that the height of adult women in a certain country is approximately normally distributed with a mean of 160 cm (5'3") and a standard deviation of 7 cm.

  • What does this mean?

    • Most women’s heights will be clustered around 160 cm.
    • Heights further away from 160 cm (both taller and shorter) become less and less likely.
    • The standard deviation of 7 cm tells us how spread out the heights are. A larger standard deviation would mean heights are more varied.

  • Using the Normal Distribution, we can answer questions like:

    • What is the probability that a randomly selected woman is taller than 170 cm? (We can calculate the area under the normal curve to the right of 170 cm).
    • What is the range of heights that contains the middle 95% of women? (We can find the interval around the mean that captures 95% of the area under the curve).