Thursday, October 18, 2012

CORRELATION AND SIMPLE LINEAR REGRESSION



IQ
TEST SCORE
X1*X2
IQ2
TS2
106
16
1696
11236
256
118
21
2478
13924
441
80
10
800
6400
100
120
25
3000
14400
625
117
16
1872
13689
256
94
11
1034
8836
121
106
6
636
11236
36
119
32
3808
14161
1024
107
20
2140
11449
400
128
15
1920
16384
225
122
26
3172
14884
676
116
18
2088
13456
324
1333
216
24644
150055
4484

Test 2 Self test I











pearson product-moment coefficient of correlation
rxy= 1
Rxy = 12(24644)-[( 1333)(216)]/ (12*150055)-(1333)2] [(12*4484)-(216)2]
     = 0.598215
The above result indicates tha there is a moderate  direct relationship between I Q and methemathical test Score . This implies hat as IQ increases Mathematical  moderately would increase  also.

B1 -           =  24644-(1333)(216)____   = 0.328
                                                    150055- (1333)2/12

Bo = Mean Y – bi meanX =   18- (0.328)(111.0833)  =  - 18.4499

The regression equation is                           
Mathematical Test Score  = -18.4499 + 0.328131 IQ


Se = Sy   = 7.36 x .80134   = 5.897
Approximately 68% of the samples will have a actual score that lie within the Se of their predicted scores YỸ +/- 5.897 = 18 +/- (5.897) . in other words 68% have scores within the range 12.103 to 23.897. Like wise, approximately 95% of the samples will have the actual scores that lie within two Se’s their predicted scores;18 +/- 11.794 or 6.206 to 29.794.

The regression equation is                           
Mathematical = -18.4499 + 0.328131 IQ                
                                                     
S = 6.18640      R-Sq = 35.8 %      R-Sq(adj) = 29.4 %

Analysis of Variance

Source            DF         SS         MS         F      P
Regression         1    213.285    213.285   5.57295  0.040
Error             10    382.715     38.271                
Total             11    596.000                           



SELF Test 1
No 1
sample
Gen Ave
Applitude
X1*X2
x12
x22
1
93
94
8742
8649
8836
2
88
82
7216
7744
6724
3
87
90
7830
7569
8100
4
89
87
7743
7921
7569
5
84
86
7224
7056
7396
6
82
85
6970
6724
7225
7
80
76
6080
6400
5776
8
91
86
7826
8281
7396
9
87
80
6960
7569
6400
10
86
81
6966
7396
6561
11
81
84
6804
6561
7056
12
88
80
7040
7744
6400
13
90
85
7650
8100
7225
14
82
84
6888
6724
7056
15
83
88
7304
6889
7744
total
1291
1268
109243
111327
107464

pearson product-moment coefficient of correlation
rxy= 1
Rxy = 15(109243)-[( 1291)(1268)]/ (15*111327)-(1291)2] [(15*107464)-(1268)2]
     = 0.453769

The above result indicates tha there is a moderate  direct relationship between General Average  and Applitude  test Score . This implies hat as General Average increases Applitude test score   moderately would increase  also.
Pearson correlation of C5 and C6 = 0.454
P-Value = 0.089



To madam Molina

a.X1 sex and X2 eduactional attainment
correlation used: Phi Coefficient

a
4

bc
6
b
3

ad
24
c
2

a+c
6
d
6

b+d
9



a+b
7



c+d
8

.rq =(6)-(24)
       __________    =   - 0.32733
           √3024

The relationship between sex and and educational attainment have definite but small negative correlation.

b. X1 sex and X3 type of employement

correlation used: Phi Coefficient

a
4

bc
16
b
4

ad
12
c
4

a+c
8
d
3

b+d
7



a+b
8



c+d
7


rq=(16)-(12)
       __________    =    0.0714
           √3136
The  result suggest that there is slight relationship between sex and type of employement



c. X2 Educational attainment and x3 Type of employement

correlation used: Phi Coefficient


a
4

bc
12
b
4

ad
16
c
3

a+c
7
d
4

b+d
8



a+b
8



c+d
7



Rq=  (12)-(16)
       __________    =    -0.07143
          √3136


The  result suggest that there is slight negative  relationship between type of employement and educational attainement

d. X2  educational attainment and x4 Rank in laedership qualities
correlation used: Rank biseral coefficient
sum with masters degree= 53 mean = 6.625
sum with out masters = 67 mean= 9.571
N= 15

Rs = 2/15 (6.625-9.571)
     = -0.3928
The realtionship between educational attainment and leadearship qualitiies have slight but not definite or can be also considered as aweak relationship.


e. X2 educational attainment and x6  verbal applitude
correlation used: point biseral coefficient
mean w1= 86.5            N1= 8
mean x6= 84.6             No= 7
SD of x6= 6.684          N = 15

.rpb      = 86.5-84.6/6.684 [√8(15)/7(14)
             = 0.314
A slight but not definite relationship is being considered in terms of the relationship between eduactional attainment and verbal applitude

f. x4 rank in leadership qualities and x5 rank in creativity test

correlation used: spearman rho correlation coefficient

x4
x5

X4-x5
(x4-x5)2


9
6

3
9

696
13
7

-4
16

225
1
5

-4
16

3360
12
14

-2
4

0.207143
7
10

-3
9

0.792857
8
8

0
0


2
4

-2
4


11
13

-2
4


15
11

4
16


10
9

1
1


4
1

3
9


5
3

2
4


14
12

2
4


13
15

-2
4


6
2

4
16






116


Rs = 1- 6 Ƹ(116)/15(152-1)
     = 0.793

The relationship between the two ranks are substantial we can say as their leadership increase their creativity also increase



G X4 rank in leadership qualities  and X6 verbal applitude
correlation used: spearman rho correlation coefficient


x4
x6
x6 ranked
x-y
(x-y)2

9
90
12.5
-3.5
12.25
5475
3
88
10
-7
49
225
1
85
7
-6
36
3360
12
86
8
4
16
1.629464
7
82
5
2
4
-0.62946
8
90
12.5
-4.5
20.25

2
89
11
-9
81

11
83
6
5
25

15
80
4
11
121

10
78
3
7
49

4
95
15
-11
121

5
91
14
-9
81

14
75
2
12
144

13
70
1
12
144

6
87
9
-3
9





912.5

Rs = 1- 6 Ƹ(912.5)/15(152-1)
     = -0.629

The relationship between the two ranks are substantial and inversely proportional as their leadership skill goes up their applitude substantialy goes down.


h. X5 rank in creativity and X6 verbal applitude

correlation used: spearman rho correlation coefficient

x5
x6
Rankedx6
x-y
(x-y)2

6
90
12.5
-6.5
42.25
5889
7
88
10
-3
9
225
5
85
7
-2
4
3360
14
86
8
6
36
1.752679
10
82
5
5
25
-0.75268
8
90
12.5
-4.5
20.25

4
89
11
-7
49

13
83
6
7
49

11
80
4
7
49

9
78
3
6
36

1
95
15
-14
196

3
91
14
-11
121

12
75
2
10
100

15
70
1
14
196

2
87
9
-7
49





981.5


Rs = 1- 6 Ƹ(981.5)/15(152-1)
     = -0.7526
The realtionship between rank increativity and verbal applitude are subtantially inversely proportional as the creatuvity increase the applitude decreases

.i X5 rank in creativity and X7 numerical proficiency
correlation used: spearman rho correlation coefficient


x5
X7
ranked
x-y
(x-y)2

6
89
13
-7
49
4773
7
87
11
-4
16
225
5
83
7.5
-2.5
6.25
3360
14
85
9.5
4.5
20.25
1.420536
10
90
14
-4
16
-0.42054
8
85
9.5
-1.5
2.25

4
81
5.5
-1.5
2.25

13
79
4
9
81

11
75
1
10
100

9
76
2
7
49

1
93
15
-14
196

3
88
12
-9
81

12
83
7.5
4.5
20.25

15
77
3
12
144

2
81
5.5
-3.5
12.25





795.5



Rs = 1- 6 Ƹ(795.5)/15(152-1)
     = -0.420

The corellation of ranked in creativity and numerical proficientcy are in moderate category somehow creativity have effect on numerical proficiency indirectly or inversly proportional as creativity increase numerical proficiency moderately decrease.


j. X6  verbal applitude and X7 numerical profficiency
pearson product-moment coefficient of correlation


x6
x7
x6*x7
x62
x72
90
89
8010
8100
7921
88
87
7656
7744
7569
85
83
7055
7225
6889
86
85
7310
7396
7225
82
90
7380
6724
8100
90
85
7650
8100
7225
89
81
7209
7921
6561
83
79
6557
6889
6241
80
75
6000
6400
5625
78
76
5928
6084
5776
95
93
8835
9025
8649
91
88
8008
8281
7744
75
83
6225
5625
6889
70
77
5390
4900
5929
87
81
7047
7569
 6561
1269
1252
106260
107983
104904







Rxy = 15(106260)-[( 1269)( 1252)]/ √[(15*107983)-(1269)2] [(15*104904)-(1252)2]
     = 0.678116

The above result indicates tha there is a substantial direct relationship between . X6  verbal applitude and X7 numerical profficiency. This implies hat as verbal applitude increases numerical proficiency substantially increase also.

Analysis of Variance
Source     DF        SS        MS        F        P
Factor      2     186.2      93.1     3.37    0.049
Error      27     745.8      27.6
Total      29     932.0
                                   Individual 95% CIs For Mean
                                   Based on Pooled StDev
Level       N      Mean     StDev  -+---------+---------+---------+-----
C9         10    24.900     7.078              (---------*---------)
C10        10    26.500     3.923                   (---------*--------)
C11        10    20.600     4.169  (---------*---------)






0 comments:

Post a Comment