关键词 > ECON20110/30370
ECON20110/30370 Econometrics Semester 1 2019/20 Midterm Examination
发布时间:2022-11-11
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
ECON20110/30370 Econometrics
Semester 1 2019/20 Midterm Examination
For questions 1-16 consider the following regression model that relates the traffic accident fatality rates to a range of variables. The data are for 47 US States and observations are for 1988.
fat = β0 + β1 unrate + β2 ydr + β3 mpp + β4 jail + u (1)
where fat is the traffic accident fatality rate (per 100,000 inhabitants), unrate is state unemployment rate (measured in percentage points, i.e. 2 = 2%), ydr is the percentage of young (aged 15-24) drivers (i.e. 10 = 10%) amongst the driving licence holders, mpp measures the average miles travelled per inhabitant per year in the state, jail is a dummy variable =1 if the there is a mandatory jail sentence for a drunk driving conviction in the state (and 0 otherwise).
On the next page you can find some summary statistics and the regression output for the model in equation (1).
The summary statistics for these variables are shown in the following output
> summary(fat_data[,c("mrall","unrate","yngdrv","mpp","jaild")])
fat
Min . :12 .31 1st Qu . :16 .27 Median :20 .15 Mean :20 .73 3rd Qu . :24 .77 Max . :32 .36
unrate
Min . : 2 .40 1st Qu . : 3 .95 Median : 5 .20 Mean : 5 .46 3rd Qu . : 6 .55 Max . :10 .90
ydr
Min . : 7 .314 1st Qu . :15 .592 Median :16 .237 Mean :16 .228 3rd Qu . :17 .117 Max . :22 .072
mpp
Min . : 5790 1st Qu . : 8043 Median : 8538 Mean : 8618 3rd Qu . : 9297 Max . :11812
jaild
Min . :0 .0000 1st Qu . :0 .0000 Median :0 .0000 Mean :0 .2979 3rd Qu . :1 .0000 Max . :1 .0000
The model is estimated on 47 observations using OLS reported in the following R output,
Call:
lm(formula = mrall ~ unrate + yngdrv + mpp + jaild, data = fat_data)
Residuals:
Min 1Q Median 3Q Max
-6 .3176 -2 .4686 -0 .5525 2 .0059 7 .0864
Coefficients:
Estimate Std . Error t value Pr(>|t |)
(Intercept) -1 .254e+01 5 .447e+00 -2 .302 0 .02637
unrate 8 .986e-01 2 .991e-01 3 .004 XXXX
ydr 2 .965e-01 2 .454e-01 1 .208 0 .23375
mpp 2 .675e-03 4 .834e-04 5 .533 1 .86e-06
jaild 1 .687e+00 1 .201e+00 1 .404 0 .16770
---
Signif . codes: 0 *** 0 .001 ** 0 .01 * 0 .05 . 0 .1 1
Residual standard error: 3 .597 on 42 degrees of freedom
Multiple R-squared: XXXX,Adjusted R-squared: 0 .5328
F-statistic: 14 .12 on 4 and 42 DF, p-value: 2 .212e-07
The correct interpretation: a number such as 4.9e _ 2 is written in scientific notation, it indicates 4.9 × 10 −2 = 0.049. We saw this when we looked at R output. Scientific notation is a standard tool taught as part of the School Maths curriculum; for a
further explanation of scientific notation see: https://www.national5maths.co.uk/scientific- notation/, https://www.bbc.co.uk/bitesize/guides/zxsv97h/revision/1, https://www.mathsisfun.com/numbers/scientific-notation.html. The way computers express numbers in scientific notation is with the ”e” or ”E” which you also saw in the exam, 4.9 × 10 −2 = 4.9e _ 2 = 4.9E _ 2 = 0.049.
If you incorrectly interpreted e as exp: so thought 4.9e _ 2 meant 4.9 × exp(_2) you would (incorrectly) get 4.9 × 0.135 = 0.66.
If you incorrectly ignored everything after the e then you would have interpreted 4.9e _ 2 as 4.9 you would (incorrectly) get 4.9.
1. Which of the following statements about the estimated coefficients is correct?
A. For an increase in 1 percentage point of unemployment rate (say from 4% to 5%) the fatality rate increases by 9.09 (deaths per 100,000).
B. As the percentage of young drivers drops by 1 percentage point (say from 13% to 12%) the fatality rate drops by 2.9 (deaths per 100,000).
C. For an increase in 1 percentage point of unemployment rate (say from 4% to 5%) the fatality rate increases by 0.909 (deaths per 100,000).
D. For an extra 1,000 miles travelled per person per year in a state, the fatality rate increases by 0.0027 (deaths per 100,000).
E. None of the above.
2. What is the residual sum of squares (RSS) for the above regression?
A. 12.93841
B. 608.1052
C. 543.4132
D. 20.73
E. None of the above.
3. In addition to the above information you know that the sum of square total (SST) is 1273.856. The value for Multiple R-squared has been omitted from the above regres- sion output. What is it?
A. 3.633
B. 0.336
C. 0.573
D. 0.533
E. None of the above.
4. Which of the following statements is correct?
A. Reject H0 : β2 = 0.18 versus HA : β1 0.18 at the 5% significance level.
B. Reject H0 : β2 = 0 versus HA : β0 0 at the 10% significance level.
C. Fail to reject H0 : β1 s _0.20 versus HA : β1 > _0.20 at the 10% significance level.
D. Fail to reject H0 : β3 = 0 versus HA : β3 0 at the 5% significance level.
E. None of the above.
5. You want to test the hypothesis that H0 : β2 = 0.5 versus HA : β2 < 0.5. Which of the following decision rules is correct when testing at a significance level of 1%.
A. Reject H0 if the t-statistic is larger than 2.423.
B. Reject H0 if the absolute value of the t-statistic is larger than 1.96.
C. Reject H0 if the t-statistic is larger than 1.96.
D. Reject H0 if the t-statistic is smaller than -2.423.
E. None of the above.
6. You want to test the hypothesis that H0 : β2 = 0.5 versus HA : β2 < 0.5. Calculate the t-statistic.
A. 1.208
B. -0.829
C. 0.604
D. 0.829
E. None of the above.
7. Which of the following is the best statement with respect to the p-value of H0 : β1 = 0 versus HA : β1 0?
A. 0.01 > p > 0.005
B. 0.01 > p > 0.005
C. 0.05 > p > 0.01
D. 0.10 > p > 0.05
E. None of the above.
8. Which of the following definitions for a 99% confidence interval of the regression coeffi- cient β4 is correct?
A. The probability that the confidence interval contains the true value of β4 is 99%.
B. The confidence interval for a given sample contains the true value of the re- gression coefficient 99% of the time.
C. 99% of the population distribution for β4 is contained in the confidence interval.
D. If confidence intervals were calculated for repeated samples then 99% of the intervals would contain the true value of β4 .
E. None of the above.
9. Calculate a 95% confidence interval of the regression coefficient β4 .
A. [_0.740, 4.114]
B. [_0.667, 4.041]
C. [_0.335, 3.709]
D. [_1.150, 4.524]
E. None of the above.
10. The coefficient on unrete has been found to be statistically significant. Which of the following statements is correct?
A. Changes in the unemployment rate, unrete, cause changes in the fatality rate, }et, if the errors are normally distribution with constant error variance.
B. The unemployment rate, unrete, may be correlated with relevant omitted vari- ables, such as the proportion of the population that is urban. This means that the effect of the unemployment rate, unrete, on fatalities, }et, can not be interpreted as causal.
C. In order to interpret the estimated effect of the unemployment rate, unrete, on fatalities, }et, as causal we require E(uIunrate) 0.
D. There are no circumstances under which we would be able to interpret the estimated effect of the unemployment rate, unrete, on fatalities, }et, as causal.
E. None of the above.
11. Assume that you want to conduct a joint hypothesis test that none of the explanatory variables but for the unemployment rate (unrate) are statistically significant. Which of the following is the restricted model with which to conduct the required F-test?
A. fat = β0 + β1unrate + u
B. fat = β0 + β2ydr + β3 mpp + β4jail + u
C. fat = β1unrate
D. fat = β0 + β1unrate + β4jail + u
E. None of the above.
12. Assume that you want to conduct a joint hypothesis test that none of the explanatory variables but for the unemployment rate (unrate) are statistically significant. In addition to the above information you also learn that the R2 of the restricted model is 0.171. What is the F-test?
A. 13.180
B. 14.778
C. 0.943
D. 39.617
E. None of the above.
13. Consider the case where the estimated model is
ln(fat) = γ0 + γ1 unrate + γ2 ydr + γ3 mpp + γ4 jail + u (2)
and we obtain 1 = 0.0467. Which of the following statements about the estimated coefficient is correct?
A. For an increase by 1% in the unemployment rate, the fatality rate increases by 0.0467 (deaths per 100,000).
B. For an increase in 1 percentage point of unemployment rate (say from 4% to 5%) the fatality rate increases by 4.67%.
C. For an increase in 1 percentage point of unemployment rate (say from 4% to 5%) the fatality rate increases by 0.0467 (deaths per 100,000).
D. For an increase in 1 percentage point of unemployment rate (say from 4% to 5%) the fatality rate increases by 0.0467%.
E. None of the above.
14. Which of the following statements is correct?
A. The p-value of a hypothesis test is the probability of the null hypothesis being correct.
B. In a regression you have discovered an economically important effect if the null hypothesis of no effect can be rejected.
C. A null hypothesis can be rejected if the p-value is larger than the set significance level.
D. When performing a hypothesis test the significance level should always be set to 5%.
E. none of the above.
15. Which of the following assumptions are not required for the OLS estimator of β2 to be unbiased.
A. The proportion of young drivers is different across states.
B. The variance of the error term is the same for each state.
C. The expected value of the error term, conditional on the proportion of young drivers, is zero.
D. The population model is defined by equation (1).
E. All of the above are required.
16. You receive extra information from another study: A state’s population density has been found to be positively related to fatality rates. Unemployment rates and population density are negatively correlated. Given this information, which of the following statements about the OLS estimator for β1 is correct?
A. E[βˆ1] = 0
B. E[βˆ1] = β1
C. E[βˆ1] < β1
D. E[βˆ1] > β1
E. None of the above.