You are on page 1of 14

PROBLEM 11-1

data :=
0 0 1 2 3 1 2 3 4 1 7 7 7 7

0 A := data 4 E := data corr ( A , B) = corr ( A , E) = 0 n := rows ( data) n=5 SD ( x) := stdev ( x) SD ( A) = 1.581 SD ( D) = 1.673
PROBLEM 11-2

1 B := data 5 F := data corr ( A , C) = 0 corr ( A , F) = 1

2 C := data

3 D := data

corr ( A , D) = 0

n n1 SD ( B) = 0 SD ( E) = 1.871 SUMY := 385 SD ( C) = 1.095 SD ( F) = 1.581 SUMXminusbarXsquared := 440 SUMXminusbarXYminusbarY := 568 SUMY n

SUMX := 220

SUMYminusbarYsquared := 1120 n := 11 SAMPLEMEANX := SAMPLEVARX := SAMPLEVARY := SUMX n

SAMPLEMEANY :=

SUMXminusbarXsquared n1 SUMYminusbarYsquared n1 SAMPLEVARX

SAMPLESTDEVX :=

SAMPLESTDEVY := SAMPLEMEANX = 20 SAMPLEVARX = 44

SAMPLEVARY SAMPLEMEANY = 35 SAMPLEVARY = 112 SAMPLESTDEVY = 10.583

SAMPLESTDEVX = 6.633 SAMPLECOVXY :=

SUMXminusbarXYminusbarY n1

SAMPLECOVXY SAMPLECORRXY := SAMPLESTDEVX SAMPLESTDEVY SAMPLECOVXY = 56.8 SAMPLECORRXY = 0.809


PROBLEM 11-3

SAMPLEMEANX := .760 SAMPLEMEANZ := 0686 SUMXminusbarXsquared := 769.081

SAMPLEMEANY := 1.037 n := 60

SUMXminusbarXYminusbarY := 720.535 SUMYminusbarYsquared := 1243.309 SUMZminusbarZsquared := 183.073 SUMXminusbarXZminusbarZ := 231.007 SUMYminusbarYZminusbarZ := 171.816 SAMPLEVARX := SUMXminusbarXsquared n1 SUMYminusbarYsquared n1

SAMPLEVARY :=

SAMPLEVARZ :=

SUMZminusbarZsquared n1

SAMPLEVARX = 13.035 SAMPLEVARY = 21.073 SAMPLEVARZ = 3.103 SAMPLESTDEVX := SAMPLESTDEVY := SAMPLESTDEVZ := SAMPLEVARX SAMPLEVARY SAMPLEVARZ SAMPLESTDEVY = 4.591

SAMPLESTDEVX = 3.61 SAMPLESTDEVZ = 1.762 SAMPLECOVXY := SAMPLECOVXZ := SAMPLECOVYZ :=

SUMXminusbarXYminusbarY n1 SUMXminusbarXZminusbarZ n1 SUMYminusbarYZminusbarZ n1 SAMPLECOVXZ = 3.915

SAMPLECOVXY = 12.212 SAMPLECOVYZ = 2.912

SAMPLECOVXY SAMPLECORRXY := SAMPLESTDEVX SAMPLESTDEVY SAMPLECOVXZ SAMPLECORRXZ := SAMPLESTDEVX SAMPLESTDEVZ SAMPLECOVYZ SAMPLECORRYZ := SAMPLESTDEVY SAMPLESTDEVZ

SAMPLECORRXY = 0.737 SAMPLECORRYZ = 0.36


PROBLEM 11-4

SAMPLECORRXZ = 0.616

data :=
0 0 1 2 3 50 70 80 60 1 8 7 6 7

0 A := data

1 B := data

corr ( A , B) = 0.9487
PROBLEM 11-5

r := .3102 ttest := r n 2 1r ttest = 1.903 r := .4156 r n 2 1r ttest = 2.664


PROBLEM 11-6

n := 36

2
this is less than critical t of 2.032, therefore H0 is not rejected.

ttest :=

2
this is more than the critical t-value of 2.032, therefore H0 is rejected.

data :=
0 0 1 2 3 4 5 0.4 0.36 0.42 0.31 0.33 0.34 1 20 25 16 30 35 30

0 A := data Number of data points: n := rows ( data) Standard deviation: SD ( x) := stdev ( x) mean ( A) = 0.36 SD ( A) = 0.042426
n 1

1 B := data

n=6

n n1 mean ( B) = 26 SD ( B) = 7.0711

i =0

( Ai mean ( A) ) ( Bi mean ( B) ) = 1.39


n 1

COVAB :=
i =0

( Ai mean ( A) ) ( Bi mean ( B) )
n1

mathcad cvar function uses n instead of n-1

COVAB = 0.278 intercept ( A , B) = 81.6 slope ( A , B) = 154.44


PROBLEM 11-7

corr ( A , B) = 0.927
regression equation is Y = 81.6 - 154.44*(X)

Varexp := 60.16 Coeff determination :=

Varunexp := 80.42 Varexp Varexp + Varunexp n := 60

Coeff determination = 0.4279

Corr :=

Coeff determination

Corr = 0.6542 Standarderror := Varunexp n2


unexplained variation is the [sum of all Ysubi - b0(est.) -b1(est.)*Xsubi]^2 --- squared vertical distances between observations and regression line (residuals)

Standarderror = 1.178 Sample var := Varexp + Varunexp n1

total variation is the sum of the squared distances of each dependent variable value minus the mean dependent value

Sample var = 2.383 Sample stdev := Samplevar

Sample stdev = 1.544


Problem 11-8

n := 5 df RSS := k RSS = 88 k

k := 1 RSS := 88 SSE nk1 df SSE := 3 SSE := 7.2

= 2.4

Totalvariation := RSS + SSE Sample var := Totalvariation n1 Totalvariation = 95.2

Sample var = 23.8 RSS := Totalvariation SSE RSS = 88


The F-statistic test whether all of the slope coefficients in the regression = 0.

RSS

F :=

k SSE nk1

F = 36.667

p := .00904

alpha := .05

if p < alpha , "reject" , "do not reject"


Problem 11-9

) = "reject"
X := 8

n := 12

b0 := .287

b1 := .802 Y := b0 + b1 X

b0 + b1 X = 6.129 standerrorest := 4.243 meanreturnmarket := 2.30 variancemeanreturn := 38.51

2 ( X meanreturnmarket) 1 varpredicterror := standerrorest 1 + + n ( n 1) variance meanreturn 2

varpredicterror = 20.884 alpha := .05 tcritical := 2.228 Y tcritical varpredicterror = 4.053 Y + tcritical varpredicterror = 16.311
Problem 11-10

n := 120

b0 := 6.071

b1 := .09251

X := 90

part A.

b0 + b1 X = 14.397 standerrorest := .81325 meanreturnmarket := 91.0983

Y := b0 + b1 X

variancemeanreturn := 137.3068

2 ( X meanreturnmarket) 1 varpredicterror := standerrorest 1 + + n ( n 1) variance meanreturn 2

varpredicterror = 0.667 alpha := .05 tcritical := 1.980 Y tcritical varpredicterror = 12.78 Y + tcritical varpredicterror = 16.014
part B.

X := 100 Y := b0 + b1 X Y = 15.322
2 ( X meanreturnmarket) 1 varpredicterror := standerrorest 1 + + n ( n 1) variance meanreturn 2

varpredicterror = 0.67009 Y tcritical varpredicterror = 13.701 Y + tcritical varpredicterror = 16.943


Problem 11-11

n := 9

SumXi := 81

SumXisquared := 789

SumYi := 144

SumXminusXbarsquared := 60

SumYminusYbarsquared := 144

SumXminusXbarYminusYbar := 84 Sumunexplainedvariationsquared := 26.4 meansampleX := SumXi n Sample varianceX := Sample stdevX := SumXminusXbarsquared n1

meansampleX = 9

Sample varianceX

Sample varianceX = 7.5 Sample stdevX = 2.7386 meansampleY := SumYi n

Sample varianceY :=

SumYminusYbarsquared n1

meansampleY = 16 Sample varianceY = 18 Sample stdevY = 4.2426 CovXY = 10.5 CorrXY = 0.9037 b1 = 1.4 b0 = 3.4

Sample stdevY := Sample varianceY SumXminusXbarYminusYbar CovXY := n1 CovXY CorrXY := Sample stdevX Sample stdevY b1 := CovXY Sample varianceX

b0 := meansampleY b1 meansampleX Y := b0 + b1 X
this is the regression equation

TotalvariationY := Sample varianceY n 1


2

ExplainedvariationY := CorrXY TotalvariationY UnexplainedvariationY := TotalvariationY ExplainedvariationY TotalvariationY = 144 ExplainedvariationY = 117.6 UnexplainedvariationY = 26.4 CorrXY = 0.8167 Coeff determination := ExplainedvariationY TotalvariationY
2

Coeff determination = 0.8167

Standerrest :=

UnexplainedvariationY n2

Standerrest = 1.942

Problem 11-12

X1

n := 1819

RSS := 14.246

SSE := 45.893 k := 1 b0 := .55851 b1 := .04375

TSS := 60.139

Y := .55851 .04375 X Coeff Rsquared := RSS TSS

Coeff Rsquared = 0.2369 X2 MultipleR := Coeff Rsquared

Standerrslopecoeff := .001842

Multiple R is simply the correlation

MultipleR = 0.4867 X3 Standerror := SSE n2

Standerror = 0.159 X4 Totaldf := n 1 Totaldf = 1818 X5 Regressiondf := k Regressiondf = 1 X6 Residualdf := n k 1 Residualdf = 1817 X7 MSSregress := RSS Regressiondf
Mean regression sum of squares

MSSregress = 14.246

X8 MSSsumofsquares := SSE Residualdf


Mean squared error

MSSsumofsquares = 0.025258 X9 F := MSSregress MSSsumofsquares

F = 564.03 X10 Slopecoeff := .04375 Slopestanderr := .001842 Slopecoeff tstatslope := Slopestanderr tstatslope = 23.75 X11 tcritical := 1.96
for alpha = .05 because df is so large, t approximates z For one independent variable, tstat = Squareroot of F

F = 23.749

lowerbound95percent := b1 tcritical Standerrslopecoeff lowerbound95percent = 0.04736 X12 upperbound95percent := b1 + tcritical Standerrslopecoeff upperbound95percent = 0.04014
Problem 11-13

Corrportretindexret := .824

This is also multiple R

Lower multiple R leads to a lower R^2, lower F and lower t significances; less exageration is seen on the regression sum of squares and the total sum of squares, affecting also the accuracy of the mean squared error. The slope coefficient is lower.

Problem 11-14

b0 := 3.021 tcalcb0 := 7.28

b1 := 7.062 tcalcb1 := 14.95 A


answer in percent

Coeff Rsquared := .794 Fee := 0

RETURN := b0 + b1 Fee RETURN = 3.021 Fee := 1 RETURN := b0 + b1 Fee RETURN = 4.041 n := 60 k := 1

answer in percent

tcritical := 2.0

result := if tcritical < tcalcb1 , "significant" , "notsignificant" result = "significant"


Problem 11-18

tstatslope := 7.988 tstatslope = 63.81


Problem 11-19

F := 63.81
The calculated F-value = the square of the t-statistic for the slope coefficient

b0 := .077 Y := b0 + b1 X Y = 4.207
Problem 11-21

b1 := .826

X := 5

r := .1452 r n 2 1r
2

n := 248

tcritical := 1.96

ttest :=

ttest = 2.302

result := if tcritical < ttest , "significant" , "not signifcant" result = "significant"


Problem 11-23

b0 := .0138 X := .01 Y := b0 + b1 X Y = 0.0203

b1 := .6486

You might also like