(Test Bank) Introduction to Econometrics

Watson,Stock, 2

Business

Test Bank

2nd Edition

• $15.00 $50.00

---

Description

Test Bank for Introduction to Econometrics, The downloadable file consists of 434 pages, and the total size of the file is 6.07 MB. Purchase now only for highly discounted price.

Author(s): Watson, Stock

---

---

Study Note Preview

Chapter  1 Economic  Questions  and Data 1.1 Multiple  Choice 1) Analyzing  the behavior  of unemployment  rates  across  U.S. states  in March  of 2006 is an example  of using A) time  series  data. B) panel  data. C) cross -sectional  data. D) experimental  data. Answer: C 2) Studying  inflation  in the United  States  from  1970 to 2006 is an example  of using A) randomized  controlled  experiments. B) time  series  data. C) panel  data. D) cross -sectional  data. Answer: B 3) Analyzing  the effect  of minimum  wage  changes  on teenage  employment  across  the 48 contiguous  U.S. states from  1980 to 2004 is an example  of using A) time  series  data. B) panel  data. C) having  a treatment  group  vs. a control  group,  since  only teenagers  receive  minimum  wages. D) cross -sectional  data. Answer: B 4) Panel  data A) is  also called  longitudinal  data. B) is the same  as time series  data. C) studies  a group  of people  at a point  in time. D) typically  uses control  and treatment  groups. Answer: A 5) Econometrics  can be defined  as follows  with the exception  of A) the  science  of testing  economic  theory. B) fitting  mathematical  economic  models  to real-world  data. C) a set of tools  used  for forecasting  future  values  of economic  variables. D) measuring  the height  of economists. Answer: D 6) To  provide  quantitative  answers  to policy  questions A) it is typically  sufficient  to use common  sense. B) you  should  interview  the policy  makers  involved. C) you  should  examine  empirical  evidence. D) is  typically  impossible  since  policy  questions  are not quantifiable. Answer: C 7) An  example  of a randomized  controlled  experiment  is when A) households  receive  a tax rebate  in one year but not the other. B) one  U.S. state  increases  minimum  wages  and an adjacent  state  does  not, and employment  differences  are observed. C) random  variables  are controlled  for by holding  constant  other  factors. D) some  5th graders  in a specific  elementary  school  are allowed  to use computers  at school  while  others  are not, and their  end-of-year performance  is compared  holding  constant  other  factors. Answer: D Stock/Watson  2e --CVC2  8/23/06  --Page  18) Ideal  randomized  controlled  experiments  in economics  are A) often  performed  in practice. B) often  used  by the Federal  Reserve  to study  the effects  of monetary  policy. C) useful  because  they give a definition  of a causal  effect. D) sometimes  used  by universities  to determine  who graduates  in four years  rather  than five. Answer: C 9) Most  economic  data are obtained A) through  randomized  controlled  experiments. B)by calibration  methods. C) through  textbook  examples  typically  involving  ten observation  points. D)by observing  real-world  behavior. Answer: D 10) One  of the primary  advantages  of using  econometrics  over typical  results  from  economic  theory,  is that A) it potentially  provides  you with quantitative  answers  for a policy  problem  rather  than simply  suggesting the direction  (positive/negative)  of the response. B) teaching  you how to use statistical  packages C) learning  how to invert  a 4 by 4 matrix. D) all  of the above. Answer: A 11) In  a randomized  controlled  experiment A) there  is a control  group  and a treatment  group. B) you  control  for the effect  that random  numbers  are not truly  randomly  generated C) you  control  for random  answers D) the  control  group  receives  treatment  on even  days  only. Answer: A 12) The  reason  why economists  do not use experimental  data more  frequently  is for all of the following  reasons except  that real-world  experiments A) cannot  be executed  in economics. B) with  humans  are difficult  to administer. C) are  often  unethical. D) have  flaws  relative  to ideal  randomized  controlled  experiments. Answer: A 13) The  most  frequently  used  experimental  or observational  data in econometrics  are of the following  type: A) cross -sectional  data. B) randomly  generated  data. C) time  series  data. D) panel  data. Answer: A Stock/Watson  2e --CVC2  8/23/06  --Page  214) In  the graph  below,  the vertical  axis represents  average  real GDP  growth  for 65 countries  over the period 1960-1995,  and the horizontal  axis shows  the average  trade  share  within  these  countries. This is an example  of A) cross -sectional  data. B) experimental  data. C) a time series. D) longitudinal  data. Answer: A Stock/Watson  2e --CVC2  8/23/06  --Page  315) The  accompanying  graph Is an example  of A) cross -sectional  data. B) experimental  data. C) a time series. D) longitudinal  data. Answer: A Stock/Watson  2e --CVC2  8/23/06  --Page  416) The  accompanying  graph   is an example  of A) experimental  data. B) cross -sectional  data. C) a time series. D) longitudinal  data. Answer: C 1.2 Essays 1) Give  at least three  examples  from  economics  where  each of the following  type of data can be used: cross -sectional  data,  time series  data,  and panel  data. Answer: Answers  will vary by student.  At this level  of economics,  students  most  likely  have  heard  of the following  use of cross -sectional  data:  earnings  functions,  growth  equations,  the effect  of class  size reduction  on student  performance  (in this chapter),  demand  functions  (in this chapter:  cigarette consumption);  time series:  the Phillips  curve  (in this chapter),  consumption  functions,  Okun ʹs law; panel data:  various  U.S. state  panel  studies  on road  fatalities  (in this book),  unemployment  rate and unemployment  benefits  variations,  growth  regressions  (across  states  and countries),  and crime  and abortion  (Freakonomics). Stock/Watson  2e --CVC2  8/23/06  --Page  5Chapter  2 Review  of Probability 2.1 Multiple  Choice 1) The  probability  of an outcome A) is  the number  of times  that the outcome  occurs  in the long run. B) equals  M × N, where  M is the number  of occurrences  and Nis the population  size. C) is  the proportion  of times  that the outcome  occurs  in the long run. D) equals  the sample  mean  divided  by the sample  standard  deviation. Answer: C 2) The  probability  of an event  A or B (Pr(Aor B)) to occur  equals A) Pr(A) × Pr(B). B) Pr(A) + Pr(B) if A and B are mutually  exclusive. C)Pr(A) Pr(B). D) Pr(A) + Pr(B) even  if A and B are not mutually  exclusive. Answer: B 3) The  cumulative  probability  distribution  shows  the probability A) that  a random  variable  is less than or equal  to a particular  value. B) of  two or more  events  occurring  at once. C) of  all possible  events  occurring. D) that  a random  variable  takes  on a particular  value  given  that another  event  has happened. Answer: A 4) The  expected  value  of a discrete  random  variable A) is  the outcome  that is most  likely  to occur. B) can  be found  by determining  the 50% value  in the c.d.f. C) equals  the population  median. D) is  computed  as a weighted  average  of the possible  outcome  of that random  variable,  where  the weights are the probabilities  of that outcome. Answer: D 5) Let  Y be a random  variable.  Then  var(Y) equals A) E[Y - μY)2]. B)E (Y - μY). C)E(Y -μY)2. D)E(Y - μY). Answer: C Stock/Watson  2e --CVC2  8/23/06  --Page  66) The  skewness  of the distribution  of a random  variable  Yis defined  as follows: A)E(Y3 - μY) σ2 Y B)E(Y - μY)3 C)EY3 - μ3 Yσ3 Y D)E(Y - μY)3 σ3 Y Answer: D 7) The  skewness  is most  likely  positive  for one of the following  distributions: A) The  grade  distribution  at your  college  or university. B) The  U.S. income  distribution. C) SAT  scores  in English. D) The  height  of 18 year old females  in the U.S. Answer: B 8) The  kurtosis  of a distribution  is defined  as follows: A)EY - μY4 σ4 Y B)EY4 - μ4 Yσ2 Y C)skewness var(Y) D)E[(Y - μY)4) Answer: A 9) For  a normal  distribution,  the skewness and kurtosis measures  are as follows: A) 1.96  and 4 B) 0 and 0 C) 0 and 3 D) 1 and 2 Answer: C Stock/Watson  2e --CVC2  8/23/06  --Page  710) The  conditional  distribution  of Y given  X=x, Pr(Y=yX=x), is A)Pr(Y = y) Pr(X = x). B)l i=1Pr(X = xi, Y = y). ∑ C)Pr(X = x, Y = y) Pr(Y = y) D)Pr(X = x, Y = y) Pr(X = x). Answer: D 11) The  conditional  expectation  of Y given  X, E(YX =x), is calculated  as follows: A)k i=1Yi Pr(X = xiY= y) ∑ B)EE(YX)] C)k i=1yi Pr(Y = yiX= x) ∑ D)l i=1E(YX= xi) ∑ Pr(X = xi) Answer: C 12) Two  random  variables  X and Y are independently  distributed  if all of the following  conditions  hold,  with the exception  of A) Pr(Y = yX = x) = Pr(Y = y). B) knowing  the value  of one of the variables  provides  no information  about  the other. C) if the conditional  distribution  of Ygiven  Xequals  the marginal  distribution  of Y. D)E(Y) = E[E(YX)]. Answer: D 13) The  correlation  between  X and YA) cannot  be negative  since  variances  are always  positive. B) is the covariance  squared. C) can  be calculated  by dividing  the covariance  between  Xand Yby the product  of the two standard deviations. D) is  given  by corr(X, Y) = cov(X, Y) var(X) var(Y). Answer: C 14) Two  variables  are uncorrelated  in all of the cases  below,  with the exception  of A)being independent. B) having  a zero covariance. C) σXY ≤ σ2 Xσ2 Y. D)E(Y  X) = 0. Answer: C Stock/Watson  2e --CVC2  8/23/06  --Page  815) var( aX + bY) = A)a2σ2 X+ b2σ2 Y. B)a2σ2 X+ 2abσXY + b2σ2 Y. C)σXY + μXμY. D)aσ2 X + bσ2 Y. Answer: B 16) To  standardize  a variable  you A) subtract  its mean  and divide  by its standard  deviation. B) integrate  the area below  two points  under  the normal  distribution. C) add  and subtract  1.96 times  the standard  deviation  to the variable. D) divide  it by its standard  deviation,  as long as its mean  is 1. Answer: A 17) Assume  that Y is normally  distributed  N(μ, σ2). Moving  from  the mean  (μ) 1.96 standard  deviations  to the left and 1.96 standard  deviations  to the right,  then the area under  the normal  p.d.f.  is A) 0.67 B) 0.05 C) 0.95 D) 0.33 Answer: C 18) Assume  that Y is normally  distributed  N(μ, σ2). To find Pr(c1 ≤ Y ≤ c2), where  c1 < c2 and di = ci–μ σ, you need to calculate  Pr(d1 ≤ Z ≤ d2) = A)Φ(d2) - Φ(d1) B)Φ(1.96)  - Φ(1.96) C)Φ(d2) - (1 - Φ(d1)) D) 1 - (Φ(d2) - Φ(d1)) Answer: A 19) If  variables  with a multivariate  normal  distribution  have  covariances  that equal  zero,  then A) the  correlation  will most  often  be zero,  but does  not have  to be. B) the  variables  are independent. C) you  should  use the χ2 distribution  to calculate  probabilities. D) the  marginal  distribution  of each of the variables  is no longer  normal. Answer: B 20) The  Student  t distribution  is A) the  distribution  of the sum of m squared  independent  standard  normal  random  variables. B) the  distribution  of a random  variable  with a chi-squared  distribution  with m degrees  of freedom,  divided by m. C) always  well approximated  by the standard  normal  distribution. D) the  distribution  of the ratio  of a standard  normal  random  variable,  divided  by the square  root of an independently  distributed  chi-squared  random  variable  with m degrees  of freedom  divided  by m. Answer: D Stock/Watson  2e --CVC2  8/23/06  --Page  921) When  there  are ∞ degrees  of freedom,  the t∞distribution A) can  no longer  be calculated. B) equals  the standard  normal  distribution. C) has  a bell shape  similar  to that of the normal  distribution,  but with “fatter”  tails. D) equals  the χ2 ∞distribution. Answer: B 22) The  sample  average  is a random  variable  and A) is  a single  number  and as a result  cannot  have  a distribution. B) has  a probability  distribution  called  its sampling  distribution. C) has  a probability  distribution  called  the standard  normal  distribution. D) has  a probability  distribution  that is the same  as for the Y1,..., Yni.i.d. variables. Answer: B 23) To  infer  the political  tendencies  of the students  at your  college/university,  you sample  150 of them.  Only  one of the following  is a simple  random  sample:  You A) make  sure that the proportion  of minorities  are the same  in your  sample  as in the  entire  student  body. B) call  every  fiftieth  person  in the student  directory  at 9 a.m.  If the person  does  not answer  the phone,  you pick the next name  listed,  and so on. C) go  to the main  dining  hall on campus  and interview  students  randomly  there. D) have  your  statistical  package  generate  150 random  numbers  in the range  from  1 to the total number  of students  in your  academic  institution,  and then choose  the corresponding  names  in the student  telephone directory. Answer: D 24) The  variance  of Y, σ2 Y, is given  by the following  formula: A)σ2 Y. B)σY n. C)σ2 Y n. D)σ2 Y n. Answer: C Stock/Watson  2e --CVC2  8/23/06  --Page  1025) The  mean  of the sample  average  Y, E(Y), is A)1 nμY . B)μY. C)μY n. D)σY μY for n > 30. Answer: B 26) In  econometrics,  we typically  do not rely on exact  or finite  sample  distributions  because A) we  have  approximately  an infinite  number  of observations  (think  of re-sampling). B) variables  typically  are normally  distributed. C) the  covariances  of Yi, Yj are typically  not zero. D) asymptotic  distributions  can be counted  on to provide  good  approximations  to the exact  sampling distribution  (given  the number  of observations  available  in most  cases). Answer: D 27) Consistency  for the sample  average  Y can be defined  as follows,  with the exception  of A)Y converges  in probability  to μY. B)Y has the smallest  variance  of all estimators. C)Y  pμY. D) the  probability  of Y being  in the range  μY ± c becomes  arbitrarily  close  to one as n increases  for any constant  c > 0. Answer: B 28) The  central  limit  theorem  states  that A) the  sampling  distribution  of Y-μY σY is approximately  normal. B)Y  p μY. C) the  probability  that Y is in the range  μY ± c becomes  arbitrarily  close  to one as n increases  for any constant c > 0. D) the  t distribution  converges  to the Fdistribution  for approximately  n>30. Answer: A 29) The  central  limit  theorem A) states  conditions  under  which  a variable  involving  the sum ofY1,..., Yni.i.d. variables  becomes  the standard  normal  distribution. B) postulates  that the sample  mean  Y is a consistent  estimator  of the population  mean  μY. C) only  holds  in the presence  of the law of large  numbers. D) states  conditions  under  which  a variable  involving  the sum of Y1,..., Yn i.i.d. variables  becomes  the Student  t distribution. Answer: A Stock/Watson  2e --CVC2  8/23/06  --Page  1130) The  covariance  inequality  states  that A) 0 ≤ σ2 XY ≤ 1. B)σ2 XY ≤ σ2 Xσ2 Y. C)σ2 XY- σ2 X ≤ σ2 Y. D)σ2 XY ≤ σ2 X σ2 Y. Answer: B 31)n i=1(axi + byi + c)= ∑ A)an i=1xi∑ + b n i=1yi∑ + n× c B)an i=1xi∑ + b n i=1yi∑ +  c C)ax + by + n×c D)an i=1xi∑ + b n i=1yi∑ Answer: A 32) n i=1(axi+b) ∑ A)n×a× x+ n×b B)n(a+b) C) D) Answer: A Stock/Watson  2e --CVC2  8/23/06  --Page  1233) Assume  that you assign  the following  subjective  probabilities  for your  final grade  in your  econometrics  course (the standard  GPA  scale  of 4 = A to 0 = F applies): Grade Probability A 0.20 B 0.50 C 0.20 D 0.08 F 0.02 The expected  value  is: A) 3.0 B) 3.5 C) 2.78 D) 3.25 Answer: C 34) The  mean  and variance  of a Bernoille  random  variable  are given  as A) cannot  be calculated B)np and np(1-p) C) p and p(1-p) D)p and (1- p) Answer: D 35) Consider  the following  linear  transformation  of a random  variable  y = x-μx σx where  μx is the mean  of x and σx is the standard  deviation.  Then  the expected  value  and the standard  deviation  of Y are given  as A) 0 and 1 B) 1 and 1 C) Cannot  be computed  because  Y is not a linear  function  of X D)μ σxand σx Answer: A Stock/Watson  2e --CVC2  8/23/06  --Page  132.2 Essays  and Longer  Questions 1) Think  of the situation  of rolling  two dice and let Mdenote  the sum of the number  of dots on the two dice.  (So M is a number  between  1 and 12.)  (a) In  a table,  list all of the possible  outcomes  for the random  variable  M together  with its probability distribution  and cumulative  probability  distribution.  Sketch  both distributions. (b) Calculate  the expected  value  and the standard  deviation  for M. (c) Looking  at the sketch  of the probability  distribution,  you notice  that it resembles  a normal  distribution. Should  you be able to use the standard  normal  distribution  to calculate  probabilities  of events?  Why  or why not? Answer: (a) Outcome   2   3         4  5      6     7  8      9    10    11   12 (sum  of dots) Probability     0.028     0.056     0.083     0.111     0.139     0.167     0.139     0.111     0.083     0.056     0.028 distribution Cumulative     0.028    0.083     0.167     0.278     0.417     0.583     0.722     0.833     0.912     0.972     1.000 probability distribution (b)  7.0; 2.42. (c)  You cannot  use the normal  distribution  (without  continuity  correction)  to calculate  probabilities  of events,  since  the probability  of any event  equals  zero. Stock/Watson  2e --CVC2  8/23/06  --Page  14