Vous êtes sur la page 1sur 15

Class Level Information

Class

Levels Values

growthgroup

2 high low

continent

4 africa america asia europe

Problem:4.1 -

Number of Observations Read 187


Number of Observations Used

187

Sum of
Squares Mean Square F Value Pr > F

Source

DF

Model

459.521554

65.645936

Error

179

800.076265

4.469700

14.69 <.0001

Corrected Total 186 1259.597819


R-Square Coeff Var Root MSE bmi Mean
0.364816

Source

8.414872

DF

2.114166

25.12417

Type I SS Mean Square F Value Pr > F

growthgroup

1 231.1455104

231.1455104

51.71 <.0001

continent

3 203.0376351

67.6792117

15.14 <.0001

growthgrou*continent

Source

25.3384087

8.4461362

1.89

0.1330

DF Type III SS Mean Square F Value Pr > F

growthgroup

33.1124419

33.1124419

7.41

continent

3 107.6019596

35.8673199

8.02 <.0001

growthgrou*continent

25.3384087

8.4461362

1.89

0.0071

0.1330

We use the univariate procedure to check the normality assumptions and find that it follows a normality assumption.

Moments
187 Sum Weights

187

0 Sum Observations

Mean

Std Deviation

2.07400224 Variance

4.3014853

Skewness

0.21076655 Kurtosis

1.10802117

Uncorrected SS 800.076265 Corrected SS


. Std Error Mean

Coeff Variation

800.076265
0.15166606

Basic Statistical Measures


Location
Mean

Variability

0.000000 Std Deviation

Median 0.018308 Variance


Mode

. Range

Interquartile Range

2.07400
4.30149
12.04058
2.01306

Stem
7
6
6
5
5
4
4
3
3
2
2
1
1
0
0
-0
-0
-1
-1
-2
-2
-3
-3
-4
-4

Leaf
3

#
1

Boxplot
*

1
Tests for Location:
Mu0=0

1
9

Test

Statistic

022344
88
112
5899
000122
5556678888999
00000122223
555555667778999999
000000111111111122333333333444
44443332222222222111111100
9988777666666555
443221111100000
9887655
3210
98766
4310
9765
332100
8765
----+----+----+----+----+----+

Student's t

Sign

Signed Rank S

p Value

6
20
3
4
2.5
6
13
116
18
30
26
16
15
7
4
5
4
4
6
4

Pr > |t|

0
| 1.0000
|
|
0.7700
|
|
0.9936
+-----+
|
|
*--+--*
|
|
|
|
+-----+
|
|
|
|
|
0
0

Pr >= |M|
Pr >= |S|

Quantiles (Definition 5)
Quantile

Estimate

100% Max

7.2789400

99%

6.0570000

95%

3.8073500

90%

2.2239288

75% Q3

0.9851312

50% Median

0.0183078

25% Q1

-1.0279322

10%

-2.9052617

5%

-3.9763912

1%

-4.6506900

0% Min

-4.7616400

Extreme Observations
Lowest

Highest

Value Obs

0
0

Value Obs

-4.76164

77 4.37823

169

-4.65069

157 4.43742

106

-4.60014

88 5.93157

61

-4.53853

14 6.05700

41

-4.34513

167 7.27894

118

Normal Probability Plot


7.25+
*
|
|
*
|
*
|
+
|
+++
|
*****
|
*+++
|
*+
|
**
|
++**
|
+****
1.25+
++***
|
+****
|
*****
|
*****
|
***+
|
****+
|
***+
|
+*+
|
+**
|
+***
|
+++**
|
+****
-4.75+**+**
+----+----+----+----+----+----+----+----+----+----+
-2
-1
0
+1
+2

Parameters for Normal


Distribution
Parameter Symbol Estimate
Mean

Mu

Std Dev

Sigma

0
2.074002

Goodness-of-Fit Tests for Normal Distribution


Test

Statistic

p Value

0.08541628 Pr > D

Kolmogorov-Smirnov D

<0.010

Cramer-von Mises

W-Sq 0.46421204 Pr > W-Sq <0.005

Anderson-Darling

A-Sq

2.48065436 Pr > A-Sq

<0.005

Quantiles for Normal


Distribution
Quantile
Percent Observed Estimated
1.0

-4.65069

-4.824851

5.0

-3.97639

-3.411430

10.0

-2.90526

-2.657941

25.0

-1.02793

-1.398893

50.0

0.01831

-0.000000

75.0

0.98513

1.398893

90.0

2.22393

2.657941

95.0

3.80735

3.411430

99.0

6.05700

4.824851

Problem 4.2 We carry out the same procedure by taking the log transformation of bmi and get the resultsClass Level Information
Class

Levels Values

growthgroup

2 high low

continent

4 africa america asia europe

Number of Observations Read 187


Number of Observations Used

Source

DF

187

Sum of
Squares Mean Square F Value Pr > F

Model

7 0.76112365

0.10873195

Error

179 1.25791711

0.00702747

Corrected Total 186 2.01904076

15.47 <.0001

R-Square Coeff Var Root MSE logbmi Mean


0.376973

2.604662

Source

0.083830

3.218460

DF Type I SS Mean Square F Value Pr > F

growthgroup

1 0.38328347

0.38328347

54.54 <.0001

continent

3 0.33768127

0.11256042

16.02 <.0001

growthgrou*continent

3 0.04015891

0.01338630

Source

1.90

0.1305

DF Type III SS Mean Square F Value Pr > F

growthgroup

0.05349616

0.05349616

7.61

0.0064

continent

0.17649046

0.05883015

8.37 <.0001

growthgrou*continent

0.04015891

0.01338630

1.90

0.1305

Class Level Information


Class

Levels Values

growthgroup

2 high low

continent

4 africa america asia europe

Number of Observations Read 187


Number of Observations Used

Source

DF

187

Sum of
Squares Mean Square F Value Pr > F

Model

7 0.76112365

0.10873195

Error

179 1.25791711

0.00702747

15.47 <.0001

Corrected Total 186 2.01904076

R-Square Coeff Var Root MSE logbmi Mean


0.376973

Source

2.604662

0.083830

3.218460

DF Type I SS Mean Square F Value Pr > F

growthgroup

1 0.38328347

0.38328347

54.54 <.0001

continent

3 0.33768127

0.11256042

16.02 <.0001

growthgrou*continent

Source

3 0.04015891

0.01338630

1.90

0.1305

DF Type III SS Mean Square F Value Pr > F

growthgroup

0.05349616

0.05349616

7.61

0.0064

continent

0.17649046

0.05883015

8.37 <.0001

growthgrou*continent

0.04015891

0.01338630

1.90

0.1305

Moments
N
Mean
Std Deviation
Skewness

187 Sum Weights


0 Sum Observations

0.08223743 Variance

0.006763

-0.01494 Kurtosis

0.70665066

Uncorrected SS 1.25791711 Corrected SS


Coeff Variation

187

. Std Error Mean

1.25791711
0.0060138

Basic Statistical Measures


Location
Mean

Variability

0.000000 Std Deviation

0.08224

Median 0.002097 Variance

0.00676

. Range

Mode

0.44534

Interquartile Range 0.07901

Tests for Location: Mu0=0


Test

Statistic

Student's t

Sign

p Value
0 Pr > |t|

1.0000

5.5 Pr >= |M| 0.4647


237 Pr >= |S|

Signed Rank S

Quantiles (Definition 5)
Quantile

Estimate

100% Max

0.25144196

99%

0.22040218

95%

0.14339134

90%

0.09243857

75% Q3

0.03976053

50% Median

0.00209743

25% Q1

-0.03924643

10%

-0.11917989

5%

-0.16670745

1%

-0.18740058

0% Min

-0.19389389

Extreme Observations
Lowest

Highest

Value Obs

Value Obs

-0.193894

14 0.170786

106

-0.187401

77 0.173065

52

-0.184457

167 0.214727

41

-0.183013

1 0.220402

61

-0.182337

157 0.251442

118

0.7501

Stem
24
22
20
18
16
14
12
10
8
6
4
2
0
-0
-2
-4
-6
-8
-10
-12
-14
-16
-18

Leaf
1
0
5

246613
23
581
4579
4012
14789990166789
01555662458
02246890133346689
11122233445556677700122233346689999
666320099666655543332110
998411887765543210
8532975521000
743294
732
955510
7664
420
64971
474320
----+----+----+----+----+----+----+
Multiply Stem.Leaf by 10**-2

#
1
1
1

Boxplot
0
0
0

6
2
3
4
4
14
11
17
35
24
18
13
6
3
6
4
3
5
6

0
|
|
|
|
|
+-----+
|
|
*--+--*
|
|
+-----+
|
|
|
|
|
|
0
0

Normal Probability Plot


0.25+
*
|
*
|
* +
|
+++
|
*****
|
* ++
|
*++
|
**
|
++**
|
+****
|
++***
0.03+
+****
|
******
|
****
|
****+
|
***++
|
**+
|
+*+
|
+***
|
++***
|
++ *
|
++ ***
-0.19+**+****
+----+----+----+----+----+----+----+----+----+----+
-2
-1
0
+1
+2

Parameters for Normal


Distribution
Parameter Symbol Estimate
Mean

Mu

Std Dev

Sigma

0.082237

Goodness-of-Fit Tests for Normal Distribution


Test

Statistic

Kolmogorov-Smirnov D

p Value

0.08645919 Pr > D

<0.010

Cramer-von Mises

W-Sq 0.44472321 Pr > W-Sq <0.005

Anderson-Darling

A-Sq 2.39426952 Pr > A-Sq

<0.005

Quantiles for Normal


Distribution
Quantile
Percent Observed Estimated
1.0

-0.18740

-0.191313

5.0

-0.16671

-0.135269

10.0

-0.11918

-0.105392

25.0

-0.03925

-0.055468

50.0

0.00210

-0.000000

75.0

0.03976

0.055468

90.0

0.09244

0.105392

95.0

0.14339

0.135269

99.0

0.22040

0.191313

Problem 4.3

Problem 4.4

Problem 4.5

===========================================================================================
Codelibname amruta "C:\Users\amrmad\Downloads";
/*Problem 4.1*/
proc import datafile ="C:\Users\amrmad\Downloads\amruta\hw3.4.xlsx" out=amruta.bmi;
sheet ='Both';
run;
proc glm data=amruta.bmi;
class growthgroup continent;
model bmi = growthgroup|continent;
output out= amruta.newbmi
residual = resid
run;
proc univariate plot data=amruta.newbmi;
var resid ;
histogram/normal;
run;
/*Problem 4.2*/
data amruta.transform (drop=bmi);
set amruta.bmi;
logbmi =log(bmi);
run;
proc glm data=amruta.transform;
class growthgroup continent;
model logbmi = growthgroup|continent;
output out= amruta.after
residual = resid2

run;
proc univariate plot data= amruta.after;
var resid2 ;
histogram/normal;
run;
/* problem 4.3*/
proc sgplot data=amruta.newbmi;
histogram resid;
run;
proc sgplot data=amruta.after;
histogram resid2;
run;
/*Problem 4.4 */
proc sgplot data= amruta.bmi;
vbar continent/response=bmi group= growthgroup
groupdisplay= cluster stat = mean;
run;
/*Problem 4.5*/
data combine;
set amruta.bmi;
combinedgrp= continent||growthgroup;
run;
proc sgplot data= combine;
Vbox bmi/ category= combinedgrp;
run;

============================ END =====================================================