ECO 391 Exam 3 Questions with Correct Answers Latest Update

ECO 391 Exam 3 Questions with Correct Answers Latest Update False - Answers T/F: The number of degrees of freedom a regression has, holding all else constant, will increase as the number of independent variables included in the regression increase. True - Answers T/F: If we create a 95% confidence interval for regression coefficient and the confidence interval does not contain 0 as s possible value for the coefficient then the coefficient will be statistically different from 0. True - Answers T/F: Stochastic implies a randomly occurring relationship. False - Answers T/F: Adding another independent variable (x) will always increase the standard error of the regression. False - Answers T/F: As a general rule when working with data, if an observation is missing data on some variable we should just fill in the data with our best guess as to what the value would logically be for the missing variable. 145 - Answers Suppose we run an OLS regression and estimate the following. Y= B0+B1X1+E b0= 10.00 b1=9.00 Given these estimates what would predict y to be for X=15. Strong Negative - Answers -1 to -.07 Moderate Negative - Answers -0.7 to -0.3 Weak Negative - Answers -0.3 to 0 Weak Positive - Answers 0 to 0.3 Moderate Positive - Answers 0.3 to 0.7 Strong Positive - Answers 0.7 to 1 This regression cannot be estimated due to misspecification of the model and too many independent variables and we would have a negative number of degrees of freedom. - Answers Suppose we wished to estimate the effect of being unemployed on a person's alcohol consumption. To do this we collect data on 50 people and include their demographic info, a full set of state dummies for the US and their current employment status as dependent variables. In total we have 63 independent variables. Given this info, which statement describes what we would likely find. The model for regression 1 fits the data best because the adjusted R^2 is higher. - Answers Suppose we run two regressions on the same data and using the same dependent variable. The first regression includes 8 independent variables and has an adjusted R2 of 0.7428. The second has 10 total independent variables (8 are the same as the first regression) and has an adjusted R2 of 0.7283. Which statement is most accurate? n-k-1 - Answers How to find degrees of freedom H0: equal 0 HA: not equal 0 - Answers State the null & alternative hypothesis t= B1-B0/SE compare t to 2. If t is 2, it is significantly different. If t is 2, it is not significantly different. - Answers Calculate the test statistic On average, holding all else constant, a one unit increase in the coefficient (ex: study hours) results in a (coefficient) increase in the dependent variable. - Answers Interpret coefficients 1.41-0/1.11= 1.27. 1.27 2 so, at the 95% level males & females are not significantly different - Answers Example: I assert that according to my estimates that the ACT scores of males and females are not significantly different. Why is it that I can assert this claim? Female coefficient is 1.41 and SE is 1.11 100-68.5= 31.5 31.5% of the variation is not explained - Answers Suppose I reported an R2 of 0.6850 for this regression. How much of the variation in test scores is not explained by this model? the variables in the regression vary by 3.457 on the ACT - Answers Suppose the SE of the regression is reported to be 3.457. Interpret in words E - Answers How does a regression model account for those variables that are no explicitly included? Sum of squared residuals on top is equal to sum of squared residuals on bottom. - Answers Ordinary Least Squares (OLS) False - Answers T/F: A correlation is the same thing as a regression coefficient False - Answers T/F: If after calculating the summary statistics for a dummy variable we find that it has a mean of 0.015 we should feel comfortable making a claim about the include group, the group the dummy variable represents, relative to the omitted group. Including the continuous variable age in this regression will make it impossible to asses the impact of being in a Greek organization on one's GPA because individuals in a Greek organization are all between 18 & 21. - Answers Suppose we collected data on GPA and other information from 100 males here at UK then ran the following regression. GPA= B0+B1Age+B2Greek+B3Male+B4Male+B5StudyHours+E