Reference no: EM132817622
For the five statements below, respond to each statement by writing "True" or "False." If the answer is false, give a one sentence explanation.
Question 1. T-tests, confidence intervals, and using p-values are all different methods of doing the same thing and they will give you the same conclusion for any particular problem.
Question 2. You should include an intercept in a regression; otherwise the slope is likely to come out different.
Question 3. If an independent variable varies more, keeping everything else constant, the variance or standard error of its slope estimate will tend to be larger.
Question 4. When using regression results to forecast the future value of Y, the forecast is more likely to be accurate if the data used are within the range of data originally used to estimate the regression.
Question 5. Any variable that is not statistically significant at 5% is making your adjusted R-squared smaller.
Explain what desirable properties will occur if the first 6 ordinary least squares classical assumptions hold. If you use any technical terms, explain what they mean.
Here are the Assumption 1-6:
Assumption: The dependent variable (x -axis) is linearly related to the coefficients, and the model contains the right independent variables.
Assumption: None of the independent variables have a perfect linear relationship with any of the other independent variables; if violated, its AKA multicollinearity.
Assumption: None of the independent variables ae correlated with the error term.
Assumption: The observed error terms are independent of each other, so they are uncorrelated with each other; If violated, it is known as "autocorrelation", or "serial correlation".
Assumption: The mean (average) of the error term is zero.
Assumption (homoskedasticity): The error term has a constant variance; if violated AKA "heteroskedasticity".
Assumption: The error term is normally distributed (optional).
This question concerns Type I and Type II errors in hypothesis testing for regression slope coefficients where the null hypothesis is Ho: B=0 and the alternative hypothesis is two-sided.
Two-sided:
Ho: B =0
Ha:B Not equal 0
In the two figures above, explain what each of the four colors (black, orange, yellow, blue) represent (Hint: one of the four areas represents Type I error and another represents Type II error.)
If we do a hypothesis test at a 1% error level instead of 5%, what happens to chance of Type II error? Briefly Explain.
A false, Positive outcome. We should have accepted the hypothesis and not reject.
Why is it that the researcher is typically trying to reject the null hypothesis in favor of the alternative, instead of the other way around?
Researchers want to usually measuremore than one variables for a sample, and computing descriptive statistics for that sample. The goal is draw conclusionabout that sample but to draw conclusions about the population that the sample was selected from. Researchers must use sample statistics to draw conclusions about the corresponding values.
a. Consider the following three models, which attempt to explain Gross Domestic Product.
a. GDP= β0 + β1 EMPLOYMENT+β2 SUMMER+β3 WINTER+β4 FALL+ β5 SPRING + e
b. GDP = β0 + β1 EMPLOYMENT+ β2 SUMMER+ β3 WINTER+ β4 FALL+ e
c. GDP= β1 EMPLOYMENT+ β2 SUMMER+ β3 WINTER+ β4 FALL+ e
where EMPLOYMENT is the number of people gainfully employed in the economy.
SUMMER is a dummy variable equal to 1 during the summer, 0 otherwise.
FALL is a dummy variable equal to 1 during the fall, 0 otherwise.
WINTER is a dummy variable equal to 1 during the winter, 0 otherwise.
SPRING is a dummy variable equal to 1 during the spring, 0 otherwise.
Which of the three models is the most appropriate? Why did you reject the other two?
(SALES) ^= 200,000 -14,275 UNEMPLOY+12,000 SUMMER+ 9,600 FALL+18,500 WINTER
where SALES is real U.S. quarterly retail sales, in millions of 1983 dollars.
UNEMPLOY is the average unemployment rate (in %) for each quarter.
SUMMER is a dummy variable equal to 1 during the summer, 0 otherwise.
FALL is a dummy variable equal to 1 during the fall, 0 otherwise.
WINTER is a dummy variable equal to 1 during the winter, 0 otherwise.
All the estimates presented above are significant at a 1% error level. The adjusted R2 is 0.89.
What is the intercept for (your answers should be numbers):
Spring
ii. Summer
iii. Fall
iv. Winter.
Interpretation of Results. Below are variable definitions and OLS results concerning per capita cigarette consumption in different countries. Interpret the results by focusing on the slope estimates and significance levels for each independent variable, and the overall goodness-of-fit of the regression. The more you can successfully explain about the results without using econometric terms or jargon, the better your score.
CIG9092 = the average number of cigarettes consumed per capita in each country, for those aged 15 and over, for 1990 to 1992.
VIOLENT CRIMES = the number of violent crimes committed per 100,000 people for each country in 1990, where murder, rape, serious assault, and robbery with violent theft are considered violent crimes.
COST = the average cost, for 1990-92, in each country for a pack of 20 cigarettes in terms of minutes of labor, including taxes.
LITERACY = the percentage of each country's population over 15 that is literate, for 1990.
GDP PER CAPITA = Gross Domestic Product per capita, 1991.
Table 1
Regression Results for 46 Countries.
Variables
|
Coefficient Estimates
|
P-Value
|
Intercept or Constant
|
3239.15
|
0.007
|
VIOLENT CRIMES
|
0.29
|
0.032
|
COST
|
-2.73
|
0.065
|
LITERACY
|
-16.15
|
0.003
|
GDP PER CAPITA
|
-0.03
|
0.006
|
N = 46.
Adjusted R2= 0.90
Attachment:- Practice Exam Questions.rar