马上要考试了，所以现在在复习计量经济学，所以小妹有几道选择题拿不准并且很疑惑，一段时间在怀疑是不是答案错了～～跪求大侠帮忙～
1   Consider the regression output for theestimated relationship between log house prices (in Euros) and square feet(x2), research university (x3), metropolitan (x4), and age (x5).  University and metropolitan are binaryvariables taking values 1 (if located in a university town and metropolitanarea, respectively) and 0 otherwise.  Amultiple regression model gives the following results for an analysis of2 000 observations 2.5 points for correct answer:

b1	b2	b3	b4	b5
0.189	0.458	0.098	0.299	0.009
(0.110)	(0.400)	(0.008)	(0.123)	(0.003)

Standard errorsbetween parentheses.

A business magazineasks for help with the interpretation of main results of these strange figures,and since you have been working with quantitative method and are considered tobe the closest export, your interpretation will be cited in the news next day:

(a)   (a)If the house size increases by 1square feet, the price will increase by 0.45%.

(b)   (b)Ifthe house size increases by 1 %, the price will increase by 0.45%.

(c)   (c) Ifthe house size increases by 1 square feet, the price will increase by 0.45% fora house located in a metropolitan area with at least one research university,controlling for age.

(d)   (d)The house prices seem to be moreclosely linked to location characteristics and age of the house compared to thesize of the house.

bibi答案是 d， 可是想不明白，看起来 最强的连接应该是size of house阿

2  18. Suppose that wewished to evaluate the factors that affected the wage income of all people inthe age of 29 and 65 in France, using national wage statistics. Which class of model would be mostappropriate?

(a)   （a)A tobit model

(b)   (b)Aninstrumental variables model

(c)    (c)Anon-stationary macro model

(d)   (d)A random-effects model

这道题好拿不准 好像被迷惑了

w

19. Which are thefollowing disadvantages of the use of pure cross-sectionaldata over panel data?

(a)   (a)The use of cross-sectional data doesnot allow the average value of the dependent variable to vary for theindividual observations

(b)   (b)Theuse of cross-sectional data does not allow for correlation between the errorterm and the explanatory variables

(c)    (c)Theuse of cross-sectional data does not enable the researcher to account for fixedeffects

(d)   (d)The use of cross-sectional does notenable the researcher to account for endogeneity

不知道选择 c 对不对

20.  How does including an irrelevant variable in aregression model affect the estimated coefficient of other variables in themodel?

(a)  They are biased downwardand have smaller standard errors

(b)  They are biased upwardand have larger standard errors

(c)  They are biased and thebias can be negative or positive

(d).  They are unbiased buthave larger standard errors

这个我敢肯定的是会增大 标准差 ，但是 应该也是有  biased upward 吧

小妹没有什么悬赏分什么的，有的只有新浪爱问的分享分，唯一能回报的估计也只有 可以帮忙寻找资料了吧

不胜感激

2

2

b