Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 10

DESCRIPTIVE STATISTICS & HISTOGRAMS

AGE
Variable
C3

Percent
100

Mean
33,754

StDev
13,981

Q1
23,000

Median
27,000

Q3
44,000

Range
51,000

Confidence Interval for Mean of Age


The assumed standard deviation = 13,981
N
207

Mean
33,754

SE Mean
0,972

95% CI
(31,849; 35,659)
Relative Frequency Histogram of Age

Frequency Histogram of Age

Normal

Normal

35

Mean
StDev
N

30

16

33,75
13,98
207

14

Percent

Frequency

20
15

Mean
StDev
N

33,75
13,98
207

10
8
6

10

2
0

10

20

30

40

50

60

70

10

20

30

AGE

60

70

Normal

Mean
StDev
N

33,75
13,98
207

100

Cumulative Percent

200

150

100

50

80

60

40

20

10

20

30

40

50

60

70

10

20

30

40
AGE

Age

RHL
Variable
C14

Percent
100

Mean
18,215

StDev
1,141

Q1
17,300

Median
18,200

Q3
19,000

Range
5,200

Median

Q3

Range

Confidence Interval for Mean of RHL


The assumed standard deviation = 1,141
N
207

50

Cumulative Relative Frequency Histogram of Age

Normal

40
AGE

Cumulative Frequency Histogram of Age

Cumulative Frequency

33,75
13,98
207

12

25

Mean
StDev
N

Mean
18,2150

SE Mean
0,0793

95% CI
(18,0596; 18,3704)

LHL
Variable

Percent

Mean

StDev

Q1

50

60

70

C15

100

18,223

1,140

17,400

18,100

19,100

5,000

Confidence Interval for Mean of LHL


The assumed standard deviation = 1,14
N
207

Mean
18,2230

SE Mean
0,0792

95% CI
(18,0677; 18,3783)

RHW
Variable
C16

Percent
100

Mean
8,3222

StDev
0,6413

Q1
7,8000

Median
8,3000

Q3
8,8000

Range
2,9000 rhw

Median
8,3000

Q3
8,8000

Range
2,7000

Confidence Interval for Mean of RHW


The assumed standard deviation = 0,6413
N
207

Mean
8,3222

SE Mean
0,0446

95% CI
(8,2348; 8,4096)

LHW
Variable
C17

Percent
100

Mean
8,3019

StDev
0,6406

Q1
7,7000

lhw

Confidence Interval for Mean of LHW


The assumed standard deviation = 0,6406
N
207

Mean
8,3019

SE Mean
0,0445

95% CI
(8,2146; 8,3892)

RWC
Variable
C18

Percent
100

Mean
16,789

StDev
1,373

Q1
15,600

Median
17,000

Q3
17,800

Median

Q3

Range
6,300 rwc

Confidence Interval for Mean of RHW


The assumed standard deviation = 1,373
N
207

Mean
16,7890

SE Mean
0,0954

95% CI
(16,6020; 16,9760)

LWC
Variable

Percent

Mean

StDev

Q1

Range

C19

100

16,793

1,382

15,700

17,000

17,800

6,800

Confidence Interval for Mean of LWC


The assumed standard deviation = 1,382
N
207

Mean
16,7930

SE Mean
0,0961

95% CI
(16,6047; 16,9813)

CHECKING ASSUMPTION
NORMALITY &EQUAL VARIANCES TESTS

normality test for Age

normality test for RHL

Normal

Normal

99,9

Mean
StDev
N
AD
P-Value

99

33,75
13,98
207
11,108
<0,005

80
70
60
50
40
30
20

99
95
90

Percent

Percent

95
90

99,9

10

0,1

0,1

20

40

60

80

14

15

16

Age

17

18
RHL

Mean
StDev
N
AD
P-Value

95
90
80
70
60
50
40
30
20

18,22
1,140
207
0,545
0,160

99
95
90

Percent

Percent

21

99,9

99

10

10

0,1

0,1

15

16

17

18
LHL

19

20

21

22

10

normality test for RHW


Normal

99,9

99,9

Mean
16,79
StDev
1,373
N
207
AD
1,436
P-Value <0,005

99
95

95
90

Percent

50
20
5

Mean
StDev
N
AD
P-Value

99

80

80
70
60
50
40
30
20
10
5

13

14

15

16

17
RWC

8,302
0,6406
207
1,109
0,006

LHW

Normal

12

Mean
StDev
N
AD
P-Value

22

80
70
60
50
40
30
20

normality test for RWC

Percent

20

Normal

Normal

0,1

19

normality test for LHW

normality test for LHL


99,9

14

18,21
1,141
207
0,560
0,146

80
70
60
50
40
30
20

10

Mean
StDev
N
AD
P-Value

18

19

20

21

0,1

9
RHW

10

8,322
0,6413
207
1,096
0,007

normality test for LWC


Normal

99,9

Mean
StDev
N
AD
P-Value

99

Percent

95
90

H0 : The distribution is distributed normal

16,79
1,382
207
1,331
<0,005

H1: The distribution is not normal

80
70
60
50
40
30
20

For tests which give a P-value smaller tahn 0.05 we


reject the hypotesis and conclude that the distribution is not
normal. So, by looking their P-values we conclude that RHL
and LHL distributes normally but the others are not.

10
5
1
0,1

12

13

14

15

16

17
LWC

18

19

20

21

ANOVA
General Linear Model: dds versus age group; bmi group; ...
Factor
age group
bmi group
job group
region
gender

Type
fixed
fixed
fixed
fixed
fixed

Levels
5
3
3
8
2

Values
1; 2; 3; 4; 5
1; 2; 3
1; 2; 3
1; 2; 3; 4; 5; 6; 7; 8
0; 1

Analysis of Variance for dds, using Adjusted SS for Tests


Source
age group
bmi group
job group
region
gender
Error
Total

DF
4
2
2
7
1
190
206

S = 6,75265

Seq SS
2531,1
3353,6
3322,3
703,9
11748,7
8663,7
30323,3

Adj SS
1949,8
569,7
369,1
146,0
11748,7
8663,7

R-Sq = 71,43%

H0 : i = 0
H1: i

Type
fixed
fixed
fixed
fixed

F
10,69
6,25
4,05
0,46
257,66

P
0,000
0,002
0,019
0,864
0,000

R-Sq(adj) = 69,02%

The only P-value grater than 0.05 is for region. So, it results in not rejecting the null hypotesis and
concluding that the for region is zero and changes in region has no significant effect on grip
strength. While other factors have significant effects on DDS.

0 for all i

en
Factor
age group
bmi group
job group
gender

Adj MS
487,5
284,8
184,6
20,9
11748,7
45,6

Levels
5
3
3
2

Values
1; 2; 3; 4; 5
1; 2; 3
1; 2; 3
0; 1

Analysis of Variance for dds, using Adjusted SS for Tests


Source
age group
bmi group
job group
gender
Error
Total
S = 6,68724

DF
4
2
2
1
197
206

Seq SS
2531,1
3353,6
3322,3
12306,6
8809,7
30323,3

Adj SS
2054,9
530,3
366,4
12306,6
8809,7

R-Sq = 70,95%

Adj MS
513,7
265,1
183,2
12306,6
44,7

F
11,49
5,93
4,10
275,20

P
0,000
0,003
0,018
0,000

R-Sq(adj) = 69,62%

TUKEYS TEST
In Tukeys test, if the confidence interval contains zero, it means that these 2 levels have no
significant difference in their means of corresponding responses. Otherwise we conclude that they have
significant difference.
Tukey 95,0% Simultaneous Confidence Intervals

Response Variable dds


All Pairwise Comparisons among Levels of gender
gender = 0 subtracted from:
gender
1

Lower
15,47

Center
17,56

Upper
19,65

-+---------+---------+---------+----(----------------*-----------------)
-+---------+---------+---------+----15,6
16,8
18,0
19,2
**Here, confidence interval does not contain zero, so we conclude that they have significant
difference between the means of the corresponding responses of the two levels of gender.
Tukey Simultaneous Tests
Response Variable dds
All Pairwise Comparisons among Levels of gender
gender = 0 subtracted from:
gender
1

Difference
of Means
17,56

SE of
Difference
1,059

T-Value
16,59

Adjusted
P-Value
0,0000

Tukey 95,0% Simultaneous Confidence Intervals


Response Variable dds
All Pairwise Comparisons among Levels of age group
age group = 1 subtracted from:
age
group
2
3
4
5

Lower
-3,92
-7,07
-11,69
-17,11

Center
0,27
-2,37
-6,54
-11,69

Upper
4,457
2,329
-1,395
-6,265

-----+---------+---------+---------+(-----*-----)
(------*-----)
(-------*------)
(------*-------)
-----+---------+---------+---------+-14,0
-7,0
0,0
7,0
**Here, confidence interval contains zero for age groups 2 and 3 , so we conclude that there is no
significant difference between the means of the corresponding responses of 1-2 and 1-3; but there is a
significant difference between 1-4 and 1-5.

age group = 2

subtracted from:

age
group
3
4
5

Center
-2,64
-6,81
-11,96

Lower
-7,72
-12,19
-17,83

Upper
2,443
-1,439
-6,079

-----+---------+---------+---------+(------*------)
(------*-------)
(-------*-------)
-----+---------+---------+---------+-14,0
-7,0
0,0
7,0
**Here, confidence interval contains zero for age group 3 , so we conclude that there is no
significant difference between the means of the corresponding responses of 2-3; but there is a
significant difference between 2-4 and 2-5.
age group = 3

subtracted from:

age
group
4
5

Center
-4,173
-9,316

Lower
-9,82
-15,50

Upper
1,476
-3,134

-----+---------+---------+---------+(-------*-------)
(--------*--------)
-----+---------+---------+---------+-14,0
-7,0
0,0
7,0

**Here, confidence interval contains zero for age group 4 , so we conclude that there is no
significant difference between the means of the corresponding responses of 3-4; but there is a
significant difference between 3-5.
age group = 4

subtracted from:

age
group
5

Center
-5,143

Lower
-11,57

Upper
1,281

-----+---------+---------+---------+(---------*--------)
-----+---------+---------+---------+-14,0
-7,0
0,0
7,0

**Here, confidence interval contains zero for age group 5 , so we conclude that there is no
significant difference between the means of the corresponding responses of 4-5.

Tukey Simultaneous Tests


Response Variable dds
All Pairwise Comparisons among Levels of age group
age group = 1 subtracted from:
age
group
2
3
4
5

Difference
of Means
0,27
-2,37
-6,54
-11,69

age group = 2
age
group
3
4
5

age
group
4
5

age
group
5

Adjusted
P-Value
0,9998
0,6362
0,0052
0,0000

SE of
Difference
1,848
1,954
2,137

T-Value
-1,429
-3,487
-5,595

Adjusted
P-Value
0,6099
0,0054
0,0000

T-Value
-2,032
-4,145

Adjusted
P-Value
0,2546
0,0005

T-Value
-2,202

Adjusted
P-Value
0,1832

subtracted from:

Difference
of Means
-4,173
-9,316

age group = 4

T-Value
0,177
-1,388
-3,496
-5,929

subtracted from:

Difference
of Means
-2,64
-6,81
-11,96

age group = 3

SE of
Difference
1,522
1,709
1,872
1,971

SE of
Difference
2,053
2,247

subtracted from:

Difference
of Means
-5,143

SE of
Difference
2,335

**As we confirmed by confidence interval tests, we get the same conclusion by the P-Value tests. When
we have a P-valus less than 0,05, we reject the null hypothesis that there is no significant
difference between the means of corresponding responses of these levels and conclude that there is a
significant difference.
Tukey 95,0% Simultaneous Confidence Intervals
Response Variable dds
All Pairwise Comparisons among Levels of bmi group
bmi group = 1 subtracted from:
bmi
group
2
3

Lower
-0,4969
1,7715

Center
3,927
6,572

Upper
8,352
11,372

-+---------+---------+---------+----(-----------*------------)
(-------------*------------)
-+---------+---------+---------+----0,0
3,5
7,0
10,5

**Here, confidence interval contains zero for bmi group 2 , so we conclude that there is no
significant difference between the means of the corresponding responses of 1-2; but there is a
significant difference between 1-3.

bmi group = 2
bmi
group
3

Lower
0,04763

subtracted from:
Center
2,644

Upper
5,241

-+---------+---------+---------+----(-------*------)
-+---------+---------+---------+----0,0
3,5
7,0
10,5
**Here, confidence interval does not contain zero for bmi group 3 , so we conclude that there is
significant difference between the means of the corresponding responses of 2-3.
Tukey Simultaneous Tests
Response Variable dds
All Pairwise Comparisons among Levels of bmi group
bmi group = 1 subtracted from:

bmi
group
2
3

Difference
of Means
3,927
6,572

bmi group = 2
bmi
group
3

SE of
Difference
1,873
2,033

T-Value
2,097
3,233

Adjusted
P-Value
0,0932
0,0041

T-Value
2,405

Adjusted
P-Value
0,0449

subtracted from:

Difference
of Means
2,644

SE of
Difference
1,099

**As we confirmed by confidence interval tests, we get the same conclusion by the P-Value tests. When
we have a P-valus less than 0,05, we reject the null hypothesis that there is no significant
difference between the means of corresponding responses of these levels and conclude that there is a
significant difference.
Tukey 95,0% Simultaneous Confidence Intervals
Response Variable dds
All Pairwise Comparisons among Levels of job group
job group = 1 subtracted from:
job
group
2
3

Lower
-2,435
0,498

Center
0,5296
5,1634

Upper
3,494
9,829

-------+---------+---------+--------(--------*-------)
(-------------*------------)
-------+---------+---------+--------0,0
3,5
7,0
**Here, confidence interval contains zero for job group 2 , so we conclude that there is no
significant difference between the means of the corresponding responses of 1-2; but there is a
significant difference between 1-3.

job group = 2

subtracted from:

job
group
3

Center
4,634

Lower
0,6741

Upper
8,593

-------+---------+---------+--------(----------*-----------)
-------+---------+---------+--------0,0
3,5
7,0

**Here, confidence interval does not contain zero for job group 3 , so we conclude that there is
significant difference between the means of the corresponding responses of 2-3.
Tukey Simultaneous Tests
Response Variable dds
All Pairwise Comparisons among Levels of job group
job group = 1 subtracted from:
job
group
2
3

Difference
of Means
0,5296
5,1634

job group = 2
job
group
3

SE of
Difference
1,255
1,975

T-Value
0,4219
2,6138

Adjusted
P-Value
0,9066
0,0260

T-Value
2,764

Adjusted
P-Value
0,0171

subtracted from:

Difference
of Means
4,634

SE of
Difference
1,677

**As we confirmed by confidence interval tests, we get the same conclusion by the P-Value tests. When
we have a P-valus less than 0,05, we reject the null hypothesis that there is no significant
difference between the means of corresponding responses of these levels and conclude that there is a
significant difference.
The following terms cannot be estimated and were removed:
gender*age group*job group
gender*age group*bmi group*job group

SCATTER PLOTS FOR INDEPENDENT VARIABLES


Scatterplot of dds vs h
70

60

60

50

50
dds

dds

Scatterplot of dds vs age


70

40

40

30

30

20

20
10

10
20

30

40

50

60

150

70

160

170

70

70

60

60

50

50

dds

dds

190

200

Scatterplot of dds vs rhl

Scatterplot of dds vs w

40

40

30

30

20

20
10

10
40

50

60

70

80
w

90

100

110

15

120

16

17

18
rhl

19

20

21

Scatterplot of rwc vs rhw

Scatterplot of dds vs rhw


70

20

60

19

50

18
rwc

dds

180
h

age

40

17
16

30

15

20

14
13

10
7,0

7,5

8,0

8,5
rhw

9,0

9,5

10,0

7,0

7,5

8,0

8,5
rhw

9,0

9,5

10,0

CORRELATION MATRIX
Correlations: rhw; rhl; rwc; age; h; w for females
rhw
0,450
0,000

rhl

rwc

0,493
0,000

0,339
0,002

age

0,269
0,017

0,049
0,669

0,541
0,000

-0,109
0,337

0,244
0,030

-0,328
0,003

-0,567
0,000

0,440
0,000

0,264
0,019

0,800
0,000

0,494
0,000

rhl

rwc

age

H0: p = 0

Cell Contents: Pearson correlation


P-Value

versus H1: p 0

where r is the correlation between a pair of variables. When


P-value is smaller than 0,05, we reject the null hypothesis. As
we see from this matrix; there is correlation between age-rhl,
height-rhw,weight-height of females ;but there is no
correlation between others.
-0,174
0,126

Correlations: rhl; rhw; rwc; age; h; w of males


rhl
0,555
0,000

rhw

rwc

0,317
0,000

0,675
0,000

age

-0,114
0,201

0,087
0,329

0,207
0,019

0,546
0,000

0,267
0,002

0,094
0,293

-0,493
0,000

0,232
0,008

0,381
0,000

0,568
0,000

-0,027
0,763

rhw

rwc

age

H0: p = 0

versus H1: p 0

where r is the correlation between a pair of variables. When


P-value is smaller than 0,05, we reject the null hypothesis. As
we see from this matrix; there is correlation between age-rhl,
age-rhw, height-rwc, height-weight of males ;but there is no
correlation between others.
0,369
0,000

Cell Contents: Pearson correlation


P-Value
Regression Analysis: dds versus age; h; w; rhl; rhw; rwc; gnd
The regression equation is
dds = - 47,9 - 0,165 age + 0,164 h + 0,0680 w + 0,392 rhl + 2,90 rhw + 1,24 rwc
+ 10,4 gnd
Predictor
Constant
age
h
w
rhl
rhw
rwc
gnd

Coef
-47,85
-0,16470
0,16380
0,06796
0,3919
2,902
1,2357
10,403

S = 6,39805

SE Coef
15,91
0,04073
0,09564
0,05109
0,7111
1,401
0,7110
1,630

R-Sq = 73,1%

T
-3,01
-4,04
1,71
1,33
0,55
2,07
1,74
6,38

P
0,003
0,000
0,088
0,185
0,582
0,040
0,084
0,000

VIF
1,632
4,292
2,938
3,315
4,061
4,794
3,170

R-Sq(adj) = 72,2%

Analysis of Variance
Source
Regression
Residual Error
Total
Source
age
h
w
rhl
rhw
rwc
gnd

DF
1
1
1
1
1
1
1

DF
7
199
206

SS
22177,2
8146,1
30323,3

MS
3168,2
40,9

F
77,40

P
0,000

Seq SS
1184,7
14877,4
1632,0
840,2
1487,1
488,1
1667,8

Factorial Fit: dds versus rhl; rhw; rwc; age; w; h; gnd


* NOTE * This design has some botched runs.
regression approach.

It will be analyzed using a

Estimated Effects and Coefficients for dds (coded units)


Term
Constant
rhl
rhw
rwc
age
w
h
gnd
rhl*rhw
rhl*rwc
rhl*age
rhl*w
rhl*h
rhl*gnd
rhw*rwc
rhw*age
rhw*w
rhw*h
rhw*gnd
rwc*age
rwc*w
rwc*h
rwc*gnd
age*w
age*h
age*gnd
w*h

Effect
11,25
-14,61
-75,80
-3,70
6,48
-8,76
141,24
-7,25
0,57
-0,09
0,29
0,16
-5,68
5,78
-0,26
-0,44
0,53
5,25
0,12
-0,14
0,16
-3,22
-0,02
0,04
-0,07
-0,03

Coef
561,93
5,62
-7,30
-37,90
-1,85
3,24
-4,38
70,62
-3,63
0,28
-0,04
0,15
0,08
-2,84
2,89
-0,13
-0,22
0,26
2,63
0,06
-0,07
0,08
-1,61
-0,01
0,02
-0,03
-0,01

SE Coef
312,629
23,172
49,098
26,535
1,344
1,962
2,117
69,974
1,748
1,065
0,068
0,100
0,104
2,763
1,830
0,138
0,190
0,296
4,495
0,074
0,057
0,149
2,782
0,006
0,007
0,167
0,010

T
1,80
0,24
-0,15
-1,43
-1,37
1,65
-2,07
1,01
-2,07
0,27
-0,64
1,45
0,75
-1,03
1,58
-0,93
-1,15
0,90
0,58
0,78
-1,22
0,54
-0,58
-1,99
2,87
-0,20
-1,44

P
0,074
0,809
0,882
0,155
0,171
0,100
0,040
0,314
0,039
0,791
0,525
0,149
0,454
0,306
0,116
0,352
0,253
0,372
0,560
0,436
0,223
0,592
0,564
0,049
0,005
0,844
0,153

Ho: There is no interaction


H1: There is interaction
We reject the null hypothesis and conclude that
there is a significant interaction between these two
factors when the P-value is less than 0.05. When we
look at the table we realize that P-values of rhl*rhw,
age*w, age*h is less than 0.05 so, there is significant
evidence to reject the null hypothesis.

w*gnd
h*gnd

0,43
-0,23

S = 6,21143
R-Sq = 77,35%

0,22
-0,11

0,222
0,362

0,98
-0,31

PRESS = 9620,56
R-Sq(pred) = 68,27%

0,331
0,755

R-Sq(adj) = 73,79%

Stepwise Regression: dds versus gnd; rhw; rwc; rhl; age; h; w


Alpha-to-Enter: 0,05 Alpha-to-Remove: 0,05
Response is dds on 7 predictors, with N = 207
Step
Constant

1
-47,85

2
-47,95

3
-57,18

gnd
T-Value
P-Value

10,4
6,38
0,000

10,4
6,38
0,000

10,1
6,26
0,000

rhw
T-Value
P-Value

2,9
2,07
0,040

3,2
2,45
0,015

3,0
2,35
0,020

rwc
T-Value
P-Value

1,24
1,74
0,084

1,29
1,84
0,067

1,82
3,22
0,001

rhl
T-Value
P-Value

0,39
0,55
0,582

age
T-Value
P-Value

-0,165
-4,04
0,000

-0,162
-4,02
0,000

-0,157
-3,89
0,000

h
T-Value
P-Value

0,164
1,71
0,088

0,188
2,22
0,028

0,223
2,79
0,006

w
T-Value
P-Value

0,068
1,33
0,185

0,064
1,27
0,206

S
R-Sq
R-Sq(adj)
Mallows Cp

6,40
73,14
72,19
8,0

6,39
73,09
72,29
6,3

6,40
72,88
72,20
5,9

You might also like