This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: 1) 2} 3) A recent study].r of IS shoppers showed that the
correlation between the time spent in the store and the
dollars Spent was 0.235. Using a signiﬁcance level
equal to 0.05, which ofthe following is the appropriate
null hypothesis to test whether the population
correlation is zero? a} Hozr=ﬂﬂ
b) Hﬂzp=0ﬂ
e] H9:;r={}.ﬂ
d) Hozpeﬂﬂ Which ofll'te following, is truc'.’ l. A study was recently conducted by Major
League Baseball to determine whether there is a
correlation between attendance at games and the
record of home team’s opponent. In this study,
the dependent variable would be the record of
the home team’s opponent. ll. A correlation of 0.9 indicates a weak linear
relationship between the variables.
lll. A perfect correlation between two variables will always produce a correlation coefﬁcient of+ Li}. a} I and II].
b) ll and [I].
C) III only. d) All are correct.
e) None are correct. In a linear regression model. why is .13. distributed
normally" a) Because 1”, is distributed normally. b) Because ofthe Central Limit Theorem.
c) Because moat variables in practice are usually
normally distributed. :1) Because we assume .5'r is normally distributed. e) Because [in and ii" are normally distributed, FﬁﬁE ‘l {.‘rF 10 F: ﬂuid. .1.st Mill:
u‘.oJ .111 mohﬁma
Am ermd‘u; F: 6mm 0..
W'sW. = .l.o C23 . ”4i .
WW. GD titElle} Mawcf‘iﬂto
Wm.
2:) our oleLam‘i— uppbj hurtI amta to} 4) 5) 6) 7) In a ordinary least squares estimation, “least squares”
refers to a) Minimizing the squared deviations of if from its mean, Y
b) Minimizing the squared deviations of K from its a prediction, Yk
c) Minimizing the value of R2
d) Minimizing the squared correlation coefficient, r3
e) Minimizing the squared deviations of )1 from the mean, Y
Which assumption is necessary to derive ﬁg and ,8}? a) 13(11 lX.) = ﬂu + [£le (linearity)
b) 5, ~ N (normality) c) EL?!) 2 0 d) ELSE") = 0': (homoskedasticity) and E(E!SI) : 0 (lack of serial correlation)
e) All ofthe above. In ﬁtting a Straight line to a set of data, the computed
R2 is 0.21. From this we can conclude a) The regression line slopes upward b) The error variance is not constant c) The errors are not normally distributed
d) Multicollinearity is a problem 6) None of the above If R2 = 1, which of the following must also be true? 3.) Changes in X cause changes in Y b) SSE = 0 c) All of the points fall exactly on a positively sloped
regression line (1) Exactly two of the are true e) All of the above are true FﬁGlEZ OF 10 R‘LWJA
>1 = wa , a:
" 4
55:94:14: MSIT 3000 Midten‘n III Spring 2008 A ‘ § WWmwww.A..~..~.W.w;‘,..."mm“... Figure 1 to the predicted value of Y? a) LetterA
b) LetterB
0) Letter C
d) LetterD
e) LetterE to the estimated value of the error term? a) LetterA
b) LetterB
0) Letter C
d) LetterD
e) LetterE PAGE 3 OF 10 8) In ﬁgure 1 above, which of the following corresponds 9) In ﬁgure 1 above, which of the following corresponds MSIT 3000 Midterm Ill Spring 2008 A The following applies to questions #10 through #17: A publishing company in New York is attempting to
develop a model that it can use to help predict textbook
sales (in 000s) for books it is considering for ﬁxture
publication. The marketing department has collected data
on several variables from a random sample of 15 books:
number of volumes of the book sold, number of pages per
book, number of competing books on the market,
advertising budget for the book, and the age of the author, SUMMARY OUTPUT
ANOVA
Signiﬁcance
df 35 MS F F
Regression 4 309603270 7740.0818 13.6076 0.0005
Residual 10 5688.0730 568.8073
Total 366484000
Standard Coefficients Error t Stat Pvalue Lower 95% Upper 95%
Intercept 1 25.3078 31.0821 4.0315 0.0024 —1 94.5630 56.0526
Pages X1 0.1759 0.0398 4.4229 0.0013 0.0873 0.2645
Competing Books X2 1 .5738 1.9959 0.7885 0.4487 6.0208 2.8733
Advertising Budget in 0005
X3 1.5917 0.4445 3,5812 0.0050 0.6014 2.5820
Age of Author X4 1.6137 0.6250 2.5819 0.0273 0.2211 3.0064 10) The regression equation for this model is a) I? : —125.3078 + 0.1759Xl —1.5738X2 —1.5917X3 +1.6137X4 b) Y : —125.3078 + 0.1759Xl —1.5738X2 —1.5917X3 +1.6137X4 c) Y : 31.0821+ 0.0398X1+1.9959X2 + 0.4445X3 +Oi6250X4 d) Y : —125.3078 +0.1759X1715738222 +1.5917ir; +1.6137X’4 6—)er ® 5).,(f v/wWQ— YonmLed/J W7l/Awta4iimm
a) IMMWHWW Mattel; WW: MAM/f)” W’Maizo W
4) TAM—Lug W “Mu“ FREE orig 5);. 11)If we wished to test the overall usefulness of the
above model, what would be the conclusion? a) The sample size is insufﬁcient to draw any
cenclusion. b) Since the relevant pvalue is 0.0005, we can
conclude at a signiﬁcance level of 1% that
the model is useful. G) Since the relevant pvalue is 0.0005, we can
conclude at a signiﬁcance level of 1% that
the model is not useful. d) Since the relevant p—value is 0.0024, we can
conclude at a signiﬁcance level of 1% that
the model is useful. :3) Since the relevant pvalue is 0.0024, we can
conclude at a signiﬁcance level of 1% that
the model is not useful. i2)The proportion of variation in textbook sales
that is explained by using the regression model
is a) 0.3448
b) 0.1552
0) 0.0735 d) 0.0005
e) Not enough information to determine. 13)How many textbooks were sampled? at) 4 b) 10 c) 14 d) 15 e) Not enough information to determine. 14) As our regression model currently stands, which
of the explanatory variables are signiﬁcantly
related to textbook sales at a 1% level of
signiﬁcance? at) Only Competing Books b) Only Pages, Advertising Budget and Age of
Author 6) Only Pages and Advertising Budget d) Only Competing Books and Age PAGES 0F 10 73am“mau
WildLEW
Mafia“. PM)
LoﬂuDJOJLGAG P_V°*£“"—
"5/ 0.0005 Lotta“,
Wait Ho‘aﬁpss:o
.«iéf P‘vmlmLx. Sana. 00003 ‘ 00!) we. 1* cm
”sesame
@ WAT 1/04).me
Y M Wot ("’le
R2: 56:2 _ Bo‘ibo.327e T55 3co43¥om
cits 0 ‘%‘(¢/% @ 731"“ dip. {oi4rf'7‘
*1 = iothdFH I‘M1:15 @
VmaW
Jug Pva/Lwa. A M) M WW F‘VWZ—DOL
TAM would 6.9. PW $62.9!le {1‘
@ .. ll 15)What is the correct interpretation of the eoeﬁieient on number of competing books in
print? a) For every additional book in print, textbook
sales drop by 15738 volumes. b) For every additional book in print, textbook
sales drop by about 1,574 volumes. e) For every additional book in print, textbook
sales increase by about 1,574 volumes. (1) For every additional 1.5738 books in print,
textbook sales decrease by 1000 volumes. 6) For every additional thousand volumes sold,
new authors decrease by 1.5738. 16) Which of the following is the point estimate for number of volumes sold if a book has 504
pages, 12 other books are on the market, the
ﬁrm spent $51,000 on advertising, and the
authoris 33 years old? a) 78.8893
b) 87.3045
0) 118.3856
d) 204.1968
6) 81174413 17) We would expect approximately 95% of actual observed textbook sales to be within how many thousand volumes of what is predicted by the
model? a) 23.8497
b) 87.9778
(1) 175.9555
d) 47.6994
8) Not enough information to determine, 0 X; : ‘IZSTBLWE {—
bqufswq) +—
Ksnsﬁz) 1*
x6717 {51) 1‘
Awe2533) ' 7878g93 @ 76 bewaremix. 11m; 18)In a regression of 100 average monthly purchases on customer age, family income and
family size: the estimated coefﬁcient on
customer age is —0.9705, with a
corresponding 95% conﬁdence interval of (
2.1892, 0.2483). From this we can conclude a) As a consumer gets one year older, his or
her spending increases by about O_97 dollars. b) The coefﬁcient on age is not signiﬁcantly
different from 0. c) Overall, the model is useful in predicting
monthly purchases. d) 95% ofthe variation in monthly purchases is
explained by customer age. 6) All of the above are either false or lack
sufficient information to make a
determination 19)In a regression of 100 average monthly purchases on customer age, family income and
family size, the estimated coefficient on family
income is 0.0023. The standard error of this
coefficient is 00007. From this we can
conclude a) Twentythree percent of the variation in
monthly purchases can be explained by
family income. 1)) Seven percent of the variation in monthly
purchases can be explained by family
income c) The coefficient on family income is
signiﬁcantly different from 0 with 95%
conﬁdence. d) The coefficient on family income is not
signiﬁcantly different from zero with 95%
conﬁdence. e) The pvalue on family income is 3.29. [ML AW #01350
4/ mmmm >
WVW uststat =M sht W2. = 00023
0.0007
0,5 04:5?” so W
valxu. 3196
$an .541 > 031.me
star aa Mama
CD = 3.23 20)Ir1 a regression of 100 average monthly purchases on customer age, family income and
family size, the estimated coefﬁcient on family
size is 8.723 with a pvalue of 0.2575. From
this we can conclude a) Since the pvalue is less than the 95%
critical value 1.96, the coefficient on family
size is not signiﬁcantly different from 0, b) Since the absolute value of the test statistic
of 33.87 is larger than the absolute value of
the 95% critical value of l.96, the
coefﬁcient on family size is signiﬁcantly
different from O. c) 25.75% of the variation in monthly
purchases can be explained by family size. d) Since the pvalue is less than 0.05, at 95%
conﬁdence, the model is not overall
significant, c) Since the pevaluc is more than 0.05, at 95%
conﬁdence, the coefficient on family size is
not significantly different from O. 21)The reduction in prediction error achieved by using a multiple regression model instead of the
sample mean is measured by a) The standard error of the model b) The standard error ofthe slope coefficient
c) The R: d) The F—statistic c) The SSE 22) Multicollinearity may cause a) The signs on the coefﬁcients to change
when a collinear variable is added to the
model b) The test Statistics to be underestimated c) The pvalues to be overestimated d) The standard errors of the coefficients to be
overestimated e) All of the above are true Pace: 8 or 10 W its(92w 410
’F'VW‘: N.
«(LMW oasrs’vas‘) w. {M‘AZ/Liéadj 23) The editors of a national automotive magazine recently studied 30 different automobiles sold in
the United States with the intent of seeing
whether they could develop a multiple
regression model to explain the variation in
highway miles per gallon A number of
different independent variables were collected.
Included in these were two qualitative variables described as follows: Car Type (categorical) Whether Car has ABS
Brakes (categorical) l = Four door sedan 1 = All wheel ABS ’ cars
3 = Compact thick 2 2 Rear wheel ABS
3 = no ABS 4 = Full size truck 5 = Sports car If these two variables are to be included in a
regression model, how many additional
variables will be needed? 24) In a multiple regression model, the adjusted R2 value a) Will always be greater than or equal to R2
b) Will always be less than or equal to R2 c) May be smaller or larger than R2,
depending on the data set and variables d) Will be equal to R2 if the added variable is
a dummy variable e) None of the above PAGE 9 0F '10 25)You have constructed a multiple regression
model for Red’s Pizza which explains pizza
slices consumed (QPlZ) by the independent
variables pizza price (PPIZ), soft drink price
(PSODA), income (INC) and a competition dummy variable (COMP). A new variable has LN Wm: 2; now become avallable to you: sav1ngs account
balance (SAVE). Suppose the correlation Mai WM and WONf coefﬁcient between income and savings account 01%; g “23 W W balance is r = 0.95. Should you consider adding savings account balance to yourmodel? CY"— 95) MW SAVE. JD
21) Yes; SAVE may signiﬁcantly improve the #L W W predicative ability of the model since it is MW .
highly reiated to the dependent variable. b) Yes; adding Variables to the model will (ED
always improve its performance. a I c) No; SAVE and INC are highly correlated )Mdm‘ if km,“ "Lilo[h 5 3 do: ‘1‘va
This multicollinearity will lead to confusing b) .LO 19W “9 4.“ W
and misleading results. VW MU. l 6L“ d) NO' models With ﬁve independent variables are’not efﬁcient. WM MAL MC
99 M {l 9.3:»:— 26) A regression ofsalary (in thousands) on age and
whether an individual has an MBA degree (l =
MBA, O : none) yields the following results: Coefﬁcients Standard Error t Stat Intercept 6974447813 1808029528 0385748557 gmw M614 : i 3 Q & AGE 205471344 3903041857 5259000326 MBA 3523632164 8130027454 4334096267 W Vim/9 a rims/Lu!
How is the coefficient on MBA interpreted? W MAJ‘6 M  lbs. :5
a) A} person With a: MBA earns $6974 more 3 5, 2,34 W
t an a person w1t out one. 
b) A person without an MBA earns $6974 L M 5* CW 
more than a person with an MBA.
c) A person with an MBA earns $32236 more @
than a person without one, d) A person without an MBA earns $32236
more than a person with an MBA. e) A person with an MBA earns $2054 more than a
person without one. ...
View
Full Document
 Fall '08
 Staff

Click to edit the document details