february 10th
Overview of Inferential Statistics
Inferential statistics involves drawing conclusions and making inferences about a general population based on sample data.
Example scenario: A company advertises to increase the chance of having a girl or boy.
Case Study: Gender Selection Company
Observations from two birth scenarios:
Scenario 1: 55 girls and 45 boys out of 100 births.
Expected outcome: Roughly 50 boys and 50 girls.
Interpretation: Not effective at increasing probability for girls; probability of observing 55 or more girls is about 18.4%.
Result is not statistically significant (high probability indicates outcome could easily be due to chance).
Scenario 2: 75 girls and 25 boys out of 100 births.
Interpretation: Effective at increasing chances for girls; probability of observing 75 or more girls is 0.00003 (very low).
Result is statistically significant (suggests actual effectiveness of the procedure).
Probability Calculations
Focus on statistical significance:
If probability is high (e.g. 18.4%), results cannot reject null assumption (ineffective company).
If probability is extremely low (e.g. 0.00003), suggests significant effect of the company.
Key Terminology
Probability Experiment: An action that yields specific outcomes (e.g. tossing a coin, rolling a die).
Event: A collection of specific outcomes from an experiment (notated with capital letters).
Example events: Tossing heads (A), rolling an odd number (B).
Sample Space: All possible outcomes of an experiment.
Example for a coin toss twice: {HH, HT, TH, TT}. For three kids: {BBB, BBG, BGB, BGG, GBB, GBG, GGB, GGG} with 8 possible outcomes.
Counting Principles
Fundamental Counting Principle: For independent events, total outcomes = product of outcomes for each stage.
Example: If flipping a coin twice, 2 outcomes per flip leads to 2 x 2 = 4 outcomes.
Tree Diagrams: A visual tool to represent outcomes systematically.
Example with Card Decks
Total of 52 cards: 26 red (Hearts, Diamonds) and 26 black (Clubs, Spades).
Each suit has 13 cards. Understand face cards (Jacks, Queens, Kings) for computations.
Probability Notation
Probability (P): Represents the likelihood of an event occurring.
Notation: P(A) for event A.
Valid probabilities range from 0 (impossible) to 1 (certain).
Expressed as fractions, decimals, or percentages (e.g., 16.7% chance of rolling a five).
Approaches to Calculate Probability
Classical Approach: Used when outcomes are equally likely.
Example: Rolling a die, probability of getting 5 = 1/6.
Relative Frequency Approach: Based on experimental results.
Larger sample sizes lead to results closer to theoretical probabilities (Law of Large Numbers).
Subjective Probability: Based on personal judgment or experience.
Example: Estimating probability of next person wearing a hat.
Complement of an Event
Complement of event A (notated as A') refers to outcomes where event A does not happen.
Probability of A and its complement should sum to 1.
Example: P(heart) = 13/52 and P(not heart) = 39/52; 13/52 + 39/52 = 1.
Conclusion
Understanding these concepts helps assess probabilities and infer conclusions from data.
Students should familiarize themselves with these principles for applications in real-world scenarios.