AP® Statistics Unit 3 flashcardsInference for Categorical Data: Proportions
34 cards · about 9 minutes for the whole deck
Flashcard drill
Space flips the card, arrow keys move, and keys 1 to 4 rate how well you knew it.
- Position
- 1 / 34
- Due
- 34
- Mastered
- 0 / 34
This card: New
Flip the card before you rate it.
Estimator
A statistic used to estimate a parameter, like p̂ for p. The value it gives from one sample is called a point estimate.
Topic 3.1: Estimators
Unbiased estimator
A statistic whose sampling distribution is centered at the parameter, so on average it neither overestimates nor underestimates.
Topic 3.1: Estimators
Sampling distribution of p̂
Mean p and standard deviation . It's approximately normal when np ≥ 10 and n(1 − p) ≥ 10.
Topic 3.2: Sampling Distributions for Sample Proportions
10% condition
When you sample without replacement, keep n ≤ 10% of the population so the observations are close enough to independent.
Topic 3.2: Sampling Distributions for Sample Proportions
Large counts condition
The normality check for proportions: at least 10 successes and 10 failures (expected counts for a test, observed counts for an interval).
Topic 3.2: Sampling Distributions for Sample Proportions
Confidence interval
A range of plausible values for a parameter: point estimate ± margin of error.
Topic 3.3: Constructing a Confidence Interval for a Population Proportion
One-sample z-interval for p
. Use it to estimate a population proportion from one random sample.
Topic 3.3: Constructing a Confidence Interval for a Population Proportion
Critical value z*
The number of standard errors to go out on each side for a given confidence level: 1.645 for 90%, 1.960 for 95%, 2.576 for 99%.
Topic 3.3: Constructing a Confidence Interval for a Population Proportion
Standard error (SE)
An estimate of a statistic's standard deviation that uses sample values. For p̂, it's .
Topic 3.3: Constructing a Confidence Interval for a Population Proportion
Margin of error
Critical value × standard error. It's how far the interval reaches on each side of the point estimate.
Topic 3.3: Constructing a Confidence Interval for a Population Proportion
Sample size for a margin of error
Solve for n, using p̂ = 0.5 if you have no estimate, and always round up.
Topic 3.3: Constructing a Confidence Interval for a Population Proportion
Interpreting a confidence interval
"We are C% confident that the interval from __ to __ captures the true [parameter in context]."
Topic 3.4: Justifying a Claim Based on a Confidence Interval for a Population Proportion
Confidence level
If you took many random samples and built an interval from each, about C% of those intervals would capture the true parameter.
Topic 3.4: Justifying a Claim Based on a Confidence Interval for a Population Proportion
Interval width
A higher confidence level makes the interval wider. A larger sample makes it narrower, and quadrupling n cuts the width in half.
Topic 3.4: Justifying a Claim Based on a Confidence Interval for a Population Proportion
Null hypothesis (H₀)
The claim of no change or no difference that a test starts out assuming, like H₀: p = 0.5. It always uses "=".
Topic 3.5: Setting Up a Test for a Population Proportion
Alternative hypothesis (Hₐ)
The claim you're looking for evidence for: p < p₀, p > p₀ or p ≠ p₀. Choose it before you see the data.
Topic 3.5: Setting Up a Test for a Population Proportion
p-value
Assuming H₀ is true, the probability of getting a statistic at least as extreme as the one observed, in the direction of Hₐ.
Topic 3.6: p-Values
Significance level (α)
The cutoff for how small a p-value must be to reject H₀, often 0.05. It's also the probability of a Type I error.
Topic 3.7: Carrying Out a Test for a Population Proportion
Reject vs. fail to reject H₀
If p-value ≤ α, reject H₀ and say there's convincing evidence for Hₐ. If not, fail to reject. You never "accept" or prove H₀.
Topic 3.7: Carrying Out a Test for a Population Proportion
Test statistic for one proportion
. It counts how many standard deviations p̂ is from the null value.
Topic 3.7: Carrying Out a Test for a Population Proportion
Type I error
Rejecting H₀ when it's actually true, a false alarm. Its probability is α.
Topic 3.8: Potential Errors When Performing Tests
Type II error
Failing to reject H₀ when it's actually false, a missed detection. Its probability is 1 − power.
Topic 3.8: Potential Errors When Performing Tests
Power
The probability of correctly rejecting a false H₀. It goes up with a larger n, a larger α, or a true value farther from the null value.
Topic 3.8: Potential Errors When Performing Tests
Sampling distribution of p̂₁ − p̂₂
Mean p₁ − p₂ and standard deviation , for independent random samples or a randomized experiment.
Topic 3.9: Sampling Distributions for the Difference Between Sample Proportions
Two-sample z-interval
. It estimates the difference p₁ − p₂.
Topic 3.10: Constructing a Confidence Interval for the Difference Between Two Population Proportions
0 in a difference interval
If an interval for p₁ − p₂ contains 0, "no difference" is plausible. If it's all above or all below 0, the proportions likely differ.
Topic 3.11: Justifying a Claim Based on a Confidence Interval for the Difference Between Two Population Proportions
Combined (pooled) proportion p̂c
Total successes ÷ total sample size from both groups. A two-proportion test uses it, since H₀ says p₁ = p₂.
Topic 3.12: Setting Up a Test for the Difference Between Two Population Proportions
Two-sample z-test statistic
. It uses the pooled proportion in the standard error.
Topic 3.13: Carrying Out a Test for the Difference Between Two Population Proportions
Chi-square test for homogeneity
Compares the distribution of one categorical variable across two or more populations or treatments, from separate samples or groups.
Topic 3.14: Setting Up a Chi-Square Test for Homogeneity or Independence
Chi-square test for independence
Checks for an association between two categorical variables measured on one sample from one population.
Topic 3.14: Setting Up a Chi-Square Test for Homogeneity or Independence
Chi-square distribution
Takes only positive values and is skewed right. Its shape depends on df, and it gets less skewed as df increases.
Topic 3.14: Setting Up a Chi-Square Test for Homogeneity or Independence
Expected count
: what you'd expect in a cell if H₀ were true. Every expected count should be greater than 5.
Topic 3.15: Carrying Out a Chi-Square Test for Homogeneity or Independence
Chi-square statistic (χ²)
over every cell. Bigger values mean the data are farther from what H₀ predicts.
Topic 3.15: Carrying Out a Chi-Square Test for Homogeneity or Independence
Degrees of freedom for χ²
For a two-way table, df = (number of rows − 1)(number of columns − 1).
Topic 3.15: Carrying Out a Chi-Square Test for Homogeneity or Independence