BAS FINAL EXAM 2025 QUESTIONS WITH ALL
PASSED ANSWERS!!
Visually, how can you tell their is an association between two categorical
variables? Answer - On a mosaic plot, the segments representing the levels of
y will vary between levels of x. see example on page (11/138) if need be
`Visually, how can you tell if an association exists between a y that is
quantitative and an x that is numerical (such as number ranges for salary, for
example)? Answer - Using side-by-side box plots, an association exists if the
mean/median lines for the box plots differ (13/138 for example).
Visually, how do you determine if their is an association between two
quantitative variables? Answer - looking at a scatterplot, the form of the
scatter should contain some order rather than just a cloud.
Interpretation for B1 in a simple linear regression model Answer - Two
individuals that differ in x by one unit will differ in y by b1 units. if B1 >0, the
person with the larger value of x is expected to have the larger y.
T/F: A simple linear relationship describes the relationship between x values
and the average value of y and cannot be used to predict individual values of y.
Answer - True
What do you have to add to a simple linear regression model in order to use it
to predict individual values of y? Answer - You have to add a disturbance
, (standard error) that represents the typical distance an individual value of y is
from the average value of y.
Under what condition do you have to test the assumptions of a simple linear
regression model Answer - if n<25.
T/F: Multiple regression models typically follow the normal distribution model.
Answer - True
What are the assumptions that need to be checked for a multiple regression
model? Answer - 1. Linearity : There should be no curvature in the residual
plots
2 equal spread : The spread of the points on the diagnostic plot should be
relatively even from left to right (no heteroscedasticity)
3. Normality: There should be no significant departure from a normal
distribution.
*If n < 25, the assumptions can just be checked visually.
What are the two methods for determining if outliers are influential points and
describe each. Answer - 1. Leverage: The leverage is the distance a particular
point is from the "predictor data cloud" (where both axis are x's). If a point is
significantly further away from the data cloud, then it is an influential point.
2. Deleted Studentized Residuals: The residual of a point based on a regression
line that is built without it. It this is large, then the point is pulling the line
toward itself and is therefore influential.
PASSED ANSWERS!!
Visually, how can you tell their is an association between two categorical
variables? Answer - On a mosaic plot, the segments representing the levels of
y will vary between levels of x. see example on page (11/138) if need be
`Visually, how can you tell if an association exists between a y that is
quantitative and an x that is numerical (such as number ranges for salary, for
example)? Answer - Using side-by-side box plots, an association exists if the
mean/median lines for the box plots differ (13/138 for example).
Visually, how do you determine if their is an association between two
quantitative variables? Answer - looking at a scatterplot, the form of the
scatter should contain some order rather than just a cloud.
Interpretation for B1 in a simple linear regression model Answer - Two
individuals that differ in x by one unit will differ in y by b1 units. if B1 >0, the
person with the larger value of x is expected to have the larger y.
T/F: A simple linear relationship describes the relationship between x values
and the average value of y and cannot be used to predict individual values of y.
Answer - True
What do you have to add to a simple linear regression model in order to use it
to predict individual values of y? Answer - You have to add a disturbance
, (standard error) that represents the typical distance an individual value of y is
from the average value of y.
Under what condition do you have to test the assumptions of a simple linear
regression model Answer - if n<25.
T/F: Multiple regression models typically follow the normal distribution model.
Answer - True
What are the assumptions that need to be checked for a multiple regression
model? Answer - 1. Linearity : There should be no curvature in the residual
plots
2 equal spread : The spread of the points on the diagnostic plot should be
relatively even from left to right (no heteroscedasticity)
3. Normality: There should be no significant departure from a normal
distribution.
*If n < 25, the assumptions can just be checked visually.
What are the two methods for determining if outliers are influential points and
describe each. Answer - 1. Leverage: The leverage is the distance a particular
point is from the "predictor data cloud" (where both axis are x's). If a point is
significantly further away from the data cloud, then it is an influential point.
2. Deleted Studentized Residuals: The residual of a point based on a regression
line that is built without it. It this is large, then the point is pulling the line
toward itself and is therefore influential.
