Justifying Claims: Slope of Regression Models - AP Statistics

1. Find $t^*$ for a 95% confidence interval with 18 degrees of freedom.

   $t^* = 2.101$. From $t$-table with $df = 18$ and $\alpha = 0.05$.

2. Which condition requires residuals to be normally distributed?

   Normality. Residuals follow a normal distribution pattern.

3. What is the purpose of a regression model?

   To predict the dependent variable based on independent variables. Models relationships to make predictions about outcomes.

4. What does $SE_b$ represent in $b \pm t^* \cdot SE_b$?

   Standard error of the slope estimate. Measures uncertainty in the slope estimate.

5. Find the margin of error for $b = 0.5$, $t^* = 2.064$, $SE_b = 0.1$.

   Margin of error = 0.2064. Margin of error = $t^* \times SE_b = 2.064 \times 0.1$.

6. Identify the alternative hypothesis for testing the slope.

   $H_a: \beta \neq 0$. Tests if there is a linear relationship between variables.

7. What does a significant $t$ statistic for slope indicate?

   Slope is significantly different from zero. Provides evidence against the null hypothesis of no relationship.

8. What does it mean if 0 is not in the confidence interval for slope?

   Reject $H_0$: Slope is significantly different from 0. Evidence of a statistically significant linear relationship.

9. What does $b$ represent in $b \pm t^* \cdot SE_b$?

   Sample estimate of the slope. The calculated slope from sample data.

10. Which condition must be met for the linear regression model?

    Linearity. Assumes a straight-line relationship between variables.

11. How is the $t$ statistic for slope calculated?

    $t = \frac{b}{SE_b}$. Tests if slope is significantly different from zero.

12. What does it mean if 0 is in the confidence interval for slope?

    Fail to reject $H_0$: Slope could be 0. No evidence of a significant linear relationship.

13. What is the consequence of violating linearity in regression?

    Model may not accurately represent the data. Linear model assumptions are violated.

14. State the formula for a confidence interval for a slope.

    $b \pm t^* \cdot SE_b$. Standard confidence interval formula using $t$ distribution for slope.

15. How do we interpret a 95% confidence interval of $(0.2, 0.8)$ for a slope?

    We are 95% confident the true slope is between 0.2 and 0.8. Expresses confidence about the true population slope.

16. What is the effect of increasing the confidence level on margin of error?

    Increases the margin of error. Higher confidence requires wider intervals to capture parameter.

17. What is the effect of a higher variability in data on the interval?

    Increases width of the confidence interval. Higher variability leads to greater uncertainty.

18. Which plot is used to check homoscedasticity?

    Residuals vs. fitted values plot. Checks for constant variance assumption.

19. What does a wide confidence interval indicate?

    Less precise estimate of the parameter. Larger range indicates more uncertainty.

20. How does the level of confidence affect the width of a confidence interval?

    Higher confidence level increases interval width. Higher confidence requires a wider range of values.

21. What does a narrow confidence interval indicate?

    More precise estimate of the parameter. Smaller range means more accurate estimation.

22. How is the width of a confidence interval related to sample size?

    Inversely related: larger sample size, narrower interval. Inverse relationship due to reduced sampling variability.

23. What role does the standard error of slope play in the interval?

    Determines interval width around $b$. Multiplied by $t^*$ to create the margin of error.

24. What is the impact of an outlier on the slope of a regression line?

    Can significantly alter the slope. Extreme values can drastically change the line's steepness.

25. What is the interpretation if a confidence interval includes negative values?

    Slope could be negative, indicating a negative relationship. Suggests the relationship may be inverse or decreasing.

26. Which condition requires independence of observations?

    Independence. Each observation is unrelated to others.

27. Which condition checks for equal variance in residuals?

    Homoscedasticity. Constant variance of residuals across all fitted values.

28. What is the typical confidence level used in practice?

    95%. Common standard in statistical practice.

29. How is the critical value $t^*$ determined?

    Based on degrees of freedom and confidence level. Uses $t$-table with appropriate degrees of freedom.

30. What does the slope of a regression line represent?

    Change in the response variable per unit increase in the predictor. Quantifies the linear relationship between variables.

31. Find and correct the error: 'Confidence interval is 0.5 to 0.9.'

    Correct: 'Confidence interval is (0.5, 0.9).'. Uses proper interval notation with parentheses.

32. When is it appropriate to use a $t$ distribution for confidence intervals?

    When sample size is small and population standard deviation is unknown. Standard conditions for using $t$ distribution in inference.

33. Identify the null hypothesis for testing the slope.

    $H_0: \beta = 0$. Tests if there's no linear relationship between variables.

34. How does sample size affect the width of a confidence interval?

    Larger sample size decreases interval width. More data points provide better precision.

35. What does the term 'regression' imply in statistics?

    Statistical process for estimating relationships among variables. Analyzes how variables are statistically related.

36. Which plot is used to check normality of residuals?

    Normal Q-Q plot. Assesses if residuals follow normal distribution.

37. What does $t^*$ represent in $b \pm t^* \cdot SE_b$?

    Critical value from the $t$ distribution. Based on confidence level and degrees of freedom.

Study these flashcards