Residuals - AP Statistics

1. What is the implication of heteroscedasticity in residuals?

   Model may need transformation. Unequal variance violates regression assumptions.

2. Find the residual if $y = 7$ and $\hat{y} = 7$.

   Residual = 0. Using the formula: $7 - 7 = 0$.

3. What is it called when residuals do not have constant variance?

   Heteroscedasticity. This indicates unequal variance across different prediction levels.

4. What does the presence of outliers in residuals suggest?

   Potential leverage points or anomalies. Outliers may indicate unusual data points or model problems.

5. What does a residual of zero imply about the fit at that point?

   Perfect fit for that observed value. The prediction exactly matches the observed value.

6. Find the residual for observed $y = 21$ and predicted $\hat{y} = 19$.

   Residual = 2. Using the formula: $21 - 19 = 2$.

7. What is a common graphical tool to evaluate the fit of a regression model?

   Residual plot. This graph reveals model assumptions and fit quality.

8. What does a large residual indicate about the model's prediction?

   Poor prediction accuracy. Large residuals show the model made significant prediction errors.

9. What does a residual plot help to identify in a regression model?

   Patterns indicating model inadequacy. Non-random patterns suggest the model needs improvement.

10. What is the residual for observed $y = 14$ and predicted $\hat{y} = 9$?

    Residual = 5. Using the formula: $14 - 9 = 5$.

11. Which plot is used to detect non-linearity in a model?

    Residual plot. Curved patterns in residuals reveal non-linear relationships.

12. State the formula for calculating a residual.

    Residual = Observed value - Predicted value. This formula shows the error between actual and predicted values.

13. Identify the term for a plot of residuals versus predicted values.

    Residual plot. This graph shows residuals on y-axis and predicted values on x-axis.

14. Find the residual for $y = 5$ and $\hat{y} = 9$.

    Residual = -4. Using the formula: $5 - 9 = -4$.

15. Find the residual with $y = 8$ and $\hat{y} = 10$.

    Residual = -2. Using the formula: $8 - 10 = -2$.

16. Find the residual for $y = 11$ and $\hat{y} = 14$.

    Residual = -3. Using the formula: $11 - 14 = -3$.

17. What is the sum of residuals in a least squares regression?

    Zero. This is a property of least squares regression.

18. What does a positive residual indicate about the observed value?

    Observed value is greater than predicted value. The model underestimated the actual value.

19. Which condition is violated if residuals display a cone shape?

    Constant variance (homoscedasticity). A cone shape shows variance increases or decreases with predicted values.

20. Which statistical measure do residuals help to check?

    Model adequacy. Residuals help determine if the model fits the data well.

21. Find the residual for $y = 17$ and $\hat{y} = 15$.

    Residual = 2. Using the formula: $17 - 15 = 2$.

22. What is the purpose of analyzing residuals in regression?

    To assess the fit of a regression model. Residuals reveal how well the model predicts actual values.

23. What does a negative residual indicate about the observed value?

    Observed value is less than predicted value. The model overestimated the actual value.

24. Which term describes residuals having constant variance?

    Homoscedasticity. This means residuals have equal spread across all predicted values.

25. Identify the term for the graphical representation of residuals.

    Residual plot. This visual tool helps identify patterns in model errors.

26. Identify the condition if residuals form a funnel shape.

    Heteroscedasticity. The funnel pattern shows non-constant variance.

27. What is the result when residuals are analyzed in a good model?

    Random scatter around zero. Random scatter indicates no systematic prediction errors.

28. Explain the implication of a residual plot showing a pattern.

    Model may be misspecified or missing variables. Patterns indicate the linear model is not appropriate.

29. Find the residual for observed value $y = 10$ and predicted value $\hat{y} = 8$.

    Residual = 2. Using the formula: $10 - 8 = 2$.

30. What is a residual in the context of regression analysis?

    A residual is the difference between an observed value and its predicted value. This measures how far each data point is from the regression line.

31. What does a systematic pattern in residuals indicate?

    Model inadequacy or bias. Systematic patterns reveal the model doesn't capture all relationships.

32. What is the residual for $y = 12$ and $\hat{y} = 13$?

    Residual = -1. Using the formula: $12 - 13 = -1$.

33. What is the residual for $y = 13$ and $\hat{y} = 16$?

    Residual = -3. Using the formula: $13 - 16 = -3$.

34. How does one interpret a residual value of zero?

    Observed equals predicted value. Perfect prediction with no error at that point.

35. Find the residual for observed $y = 15$ and predicted $\hat{y} = 18$.

    Residual = -3. Using the formula: $15 - 18 = -3$.

36. What does a residual plot with no discernible pattern indicate?

    A good fit for the regression model. Random scatter indicates the model captures the relationship well.

37. Which term describes residuals that display constant spread?

    Homoscedasticity. Equal spread indicates the regression assumptions are met.

38. What does a pattern in a residual plot suggest about a model?

    The model may be inadequate or misspecified. Patterns suggest the linear model doesn't fit the data properly.

39. What is indicated by residuals tightly clustering around zero?

    Good model fit. Small residuals near zero indicate accurate predictions.

Study these flashcards