Statistics and Probability

TL;DR

Statistics is about making sense of data, while probability helps us understand the likelihood of events. Together, they form a powerful toolkit for making informed decisions based on uncertain information. You'll learn how to describe data and predict outcomes using these fundamental concepts.

1. The Mental Model

Think of statistics as detectives investigating data to find patterns and stories, while probability is like a fortune teller, predicting how likely future events are based on those patterns. They're two sides of the same coin, helping you understand the world around you with numbers.

2. The Core Material

Statistics and probability help us analyze and interpret data. Let's break down the key areas.

Descriptive Statistics

This is about summarizing and organizing data so it's easier to understand. You're simply describing what you see.

• Measures of Central Tendency: These tell you where the "middle" or "average" of your data lies.
  • Mean: The traditional average (sum of all values divided by the count).
  • Median: The middle value when data is ordered. It's great for skewed data because it's not affected by extreme values (outliers).
  • Mode: The value that appears most often. Useful for categorical data.
• Measures of Spread (or Variability): These tell you how spread out your data is.
  • Range: The difference between the highest and lowest values. Simple but sensitive to outliers.
  • Variance and Standard Deviation: These measure the average distance of each data point from the mean. Standard deviation is the square root of variance and is often preferred because it's in the same units as your original data.

Probability

Probability is the chance of an event happening. It's expressed as a number between 0 (impossible) and 1 (certain), or as a percentage.

• Basic Probability Formula:
  $$ P(\text{Event}) = \frac{\text{Number of favorable outcomes}}{\text{Total number of possible outcomes}} $$
• Types of Events:
  • Independent Events: The outcome of one event doesn't affect the outcome of another (e.g., flipping a coin twice).
  • Dependent Events: The outcome of one event does affect the outcome of another (e.g., drawing two cards from a deck without replacement).
• Key Concepts:
  • Union ($ \cup $): The probability of A or B happening. $P(A \cup B) = P(A) + P(B) - P(A \cap B)$.
  • Intersection ($ \cap $): The probability of A and B happening. For independent events, $P(A \cap B) = P(A) \times P(B)$.
  • Conditional Probability: The probability of an event occurring given that another event has already occurred. $P(A|B) = \frac{P(A \cap B)}{P(B)}$.

Inferential Statistics

This is where you use data from a sample to make predictions or draw conclusions about a larger population. You're "inferring" things.

• Sampling: It's often impossible to collect data from everyone, so we take a sample. The goal is a sample that represents the population well.
• Hypothesis Testing: This is a formal process to test a claim (hypothesis) about a population using sample data.
  • You formulate a null hypothesis ($H_0$), which is usually a statement of no effect or no difference.
  • You formulate an alternative hypothesis ($H_1$), which is what you're trying to prove.
  • You collect data and perform a statistical test to see if there's enough evidence to reject the null hypothesis.
• Confidence Intervals: Instead of just a single estimate, a confidence interval gives you a range of values within which the true population parameter is likely to fall, with a certain level of confidence (e.g., 95% confidence).

Here's a simple flow of how we often approach a data question:

graph TD
    A["Have a Question/Problem"] --> B["Collect Data (Sample)"];
    B --> C{"Is the data clear?"};
    C -- "No, needs summarizing" --> D["Descriptive Statistics (Mean, Median, Std Dev)"];
    C -- "Yes, but need to predict" --> E["Probability (Likelihood of events)"];
    D --> F["Understand Data Patterns"];
    E --> F;
    F --> G{"Need to generalize?"};
    G -- "Yes, to population" --> H["Inferential Statistics (Hypothesis Tests, CIs)"];
    G -- "No, just describe" --> I["Draw Conclusions/Make Decisions"];
    H --> I;

3. Worked Example

Let's say you're a game developer, and you've launched a new game level. You collect data on how long (in minutes) 10 randomly selected players took to complete it:

[12, 15, 13, 20, 14, 12, 18, 13, 16, 17]

Task: Calculate the mean, median, and standard deviation of completion times.

1. Order the data: [12, 12, 13, 13, 14, 15, 16, 17, 18, 20] (This helps for the median).

2. Calculate the Mean:
   Sum = $12+12+13+13+14+15+16+17+18+20 = 150$
   Count = 10
   Mean = $150 / 10 = 15$ minutes.

3. Calculate the Median:
   Since there are 10 (an even number) data points, the median is the average of the two middle values (the 5th and 6th values).
   5th value = 14
   6th value = 15
   Median = $(14 + 15) / 2 = 14.5$ minutes.

4. Calculate the Standard Deviation: This is a bit more involved, but here's how you'd do it step-by-step (or use a calculator/software).

   • First, find the variance. For each data point, subtract the mean, square the result, and then sum all these squared differences. Finally, divide by (n-1) for a sample.
     Differences from mean (15):
     $(-3)^2, (-3)^2, (-2)^2, (-2)^2, (-1)^2, (0)^2, (1)^2, (2)^2, (3)^2, (5)^2$
     $9, 9, 4, 4, 1, 0, 1, 4, 9, 25$
     Sum of squared differences = $9+9+4+4+1+0+1+4+9+25 = 66$
     Variance ($s^2$) = $66 / (10-1) = 66 / 9 \approx 7.33$
   • Then, find the standard deviation ($s$) by taking the square root of the variance.
     Standard Deviation = $ \sqrt{7.33} \approx 2.71$ minutes.

Conclusion: On average, players complete the level in 15 minutes, with half completing it in 14.5 minutes or less. The completion times typically vary by about 2.71 minutes from the mean.

4. Key Takeaways

• Descriptive statistics summarize data using measures like mean, median, mode (central tendency) and range, variance, standard deviation (spread).
• Probability quantifies the chance of an event, expressed as a number between 0 and 1.
• Independent events don't affect each other, while dependent events do.
• Inferential statistics allows you to make educated guesses about a larger population based on a smaller sample.
• Hypothesis testing helps validate claims about populations using sample data.
• Confidence intervals provide a range where a true population parameter likely lies.

Common Mistakes to Avoid:
- Confusing mean and median: Use the median for skewed data or when outliers are present to get a better sense of the "typical" value.
- Misinterpreting probability: A probability of 0.5 means a 50% chance, not that it will happen exactly half the time in a small number of trials.
- Drawing conclusions from small samples: Small, unrepresentative samples can lead to wildly inaccurate inferences about a population.
- Forgetting units: Always state the units (e.g., minutes, dollars, counts) for your statistical measures to make them meaningful.

5. Now Try It

You're tracking daily visitors to your personal blog over a week: [250, 300, 275, 290, 320, 280, 260]. Calculate the mean and median number of visitors for the week. Then, imagine a new blog post makes your visitor count for the next day jump to 1500. How does this outlier affect the mean compared to the median? What does this tell you about which measure might be better for visitor data? Write down your calculations and a short explanation.