AP Stats Chapter 12 Flashcards

Scatterplots illustrate the linear relationship between a quantitative explanatory variable (x) and a quantitative response variable (y).
To predict y for a given value of x, use the least-squares regression line fitted to the data, providing estimates based on observed points.

The population regression line represents the true regression line that incorporates all observations.
The sample regression line is an estimated regression line derived from the least-squares regression line of a sample.

In a least-squares regression line, the statistic 'a' signifies the intercept, while 'b' denotes the slope.
Values of 'a' and 'b' are obtained from sample data and can change with repeated sampling; 'a' estimates the population intercept (α), and 'b' estimates the population slope (β).

LINEAR: The true relationship between x and y should be linear. Checking residuals and scatterplots can validate this.
INDEPENDENT: Observations must be independent. Use the 10% rule for sampling without replacement and ensure independence in experimental results.
NORMAL: Assess normality of residuals with stemplots, histograms, or normal probability plots for any skewness or significant deviations.
EQUAL VARIANCE: The scatter above and below the residual=0 line in the residual plot should display similar amounts of scatter.
RANDOM: Data should come from random samples or randomized experiments for integrity in analysis.

Statistics 'a' and 'b' are unbiased estimators of the population parameters α and β respectively.

Population standard deviation (σ) measures variability of y about the population regression line; it is estimated using the standard deviation of residuals.
Residuals from the least-squares regression line quantify how much y varies around the population regression line.

Estimate the spread of the sampling distribution with the standard error of the slope, calculated as SEb = s/Sx * √(n-1), where n is the sample size, and it has (n-2) degrees of freedom.

σ indicates the variability of response variable y around the population regression line.
SEb represents the spread of the sampling distribution.

The confidence interval for the population slope β is constructed as: statistic ± critical value (standard deviation of statistic), expressed as b ± t*SEb.
The test statistic for the slope of a least-squares regression line is given by t = (b - βo)/SEb.
To find the p-value for a t-test, utilize the t distribution/table with degrees of freedom (df) equal to n - 2.