Question

11) Which of the following is NOT an assumption of Simple Linear Regression?

a. There is constant variance with the residuals
b. The residuals are normally distributed
c. The independent variables are not collinear
d. The values of the residuals are independent of one another.
12) Which of one the following graph types would be best if you wanted to show the number of
occurrences (frequency) of data that are divided into categories?
a. Strip chart
b. Bar graph
c. Box plot
d. Histogram.
13) When standardizing measures of distance what do the Z-values refer to?
a. Units of distance from the mean
b. All P-values contained in the Probability Distribution Table
c. Areas to the right in the histogram
d. The range of data contained in the histogram.
14) In regression analysis if the R2 value is high, this means what?
a. There are other factors that explain or predict the dependent variable that are not
included among the independent variables
b. Y values cannot be predicted based on given X values
c. The model does not fit the data very well
d. None of the above.
15) In an OLS regression model the residual error can be defined as which of the following?
a. The difference between the regression line and the observed X
b. The difference between predicted Y and observed Y
c. The estimated ~1 value
d. The estimate of the regression.
16) If a Moran's I value equals zero this means ...
a. A dispersed pattern
b. There is negative spatial autocorrelation
c. There is no spatial autocorrelation
d. A clustered pattern.
17) A GWR model is "Miss-Specified" if...
a. It has a low Adjusted R2
b. There is randomness (dependence) in the model residuals
c. It has a high Adjusted R2
d. A single variable explains 100% of the geographic phenomena being tested.
18) A "leverage point" in an ordinary least squares regression refers to:
a. An outlier
b. An observation that is similar to its neighbours
c. A point that can influence the regression in a way that reflects the value of most
other points
d. None of the above.
19) A Type I Error means which of the following:
a. Accepting the Null hypothesis
b. Accepting the Null hypothesis incorrectly
c. Rejecting the Null hypothesis incorrectly
d. Rejecting both the Alternative and the Null hypothesis.
20) If points do not align with the diagonal lines in a QQ-Plot this tells us that ...
a. There is skewness in the data
b. The variable has a weak positive correlation
c. The variable is not normally distributed
d. None of the above.

Answer 1

11) c. the independent variables are not col-linear

12) b. Bar graph

13) a. Units of distance from the mean

14) d. None of the above

15) b. The difference between the predicted Y and observed Y

16) c. There is no spatial auto-correlation

17) a. It has a low Adjusted R²

18) c. A point that can influence the regression in a way that reflects the value of most other points

19) c. Rejecting the Null hypothesis incorrectly

20) c. The variable is not normally distributed

Explanation:

11) With simple linear regression there is a col linear relationship between the dependent variable and the independent variables

Therefore, the correct option is c. the independent variables are not col linear

12) The bar graph which is also known as a bar chart, is used for the frequencies of data that are categorized

Therefore, the correct option is b. Bar graph

13) The z-value is a measure of the distance from the mean, therefore, the correct option is a. Units of distance from the mean

14) The higher the R² value the increased correlation between the factors of the dependent variable and the independent variable

Therefore, the correct option is d. None of the above

15) In an OLS residual error = Observed value of Y - Predicted value of Y

Therefore, the best option is b. The difference between the predicted Y and observed Y

16) A Moran's I value of zero indicates perfect randomness or no auto-correlation

The correct option is therefore, c. There is no spatial auto-correlation

17) The adjusted R² value increases when the added variable improves the model to the observation correlation more than would be expected by chance alone, therefore, the correct option is a. It has a low Adjusted R²

18) A leverage point is a point that is far from the other points in the observation, therefore, the correct option is c. A point that can influence the regression in a way that reflects the value of most other points

19) A type 1 error occurs when an accurate (statistically acceptable) null hypothesis is rejected therefore, the correct option is c. Rejecting the Null hypothesis incorrectly

20) The Quantile-Quanttile (QQ) plot is a technique used for checking the commonness of the sources of two sets of data, therefore sources of data that are from the source (hence normally distributed) align with the diagonal lines in a QQ-Plot

The correct option is therefore c. The variable is not normally distributed