Obesity Prediction Model Analysis

This assignment focuses on the analysis of a logistic regression model used to predict obesity. The model incorporates variables such as the number of parents with university education ('edu_par') and the presence or absence of obese parents ('owob_par'). The output presents key model statistics, including the -2 Log likelihood, R-squared values, and classification table. It also highlights the variable coefficients and their significance in predicting obesity.

Bio statistics manuscript
TASK 1
1

TABLE OF CONTENTS
1....................................................................................................................................................................................................................2
a. Mean, standard deviation, minimum and maximum values for age ...................................................................................................2
b. Frequency (% and number) of students in each of new age categories...............................................................................................2
2. ..................................................................................................................................................................................................................3
Descriptive statistics for demographic information ................................................................................................................................3
3....................................................................................................................................................................................................................3
Mean for aggression, thrill seeking and risk acceptance differing from various demographics.............................................................3
4....................................................................................................................................................................................................................9
In terms of depression is there difference among depressed and not depressed group in accordance with following............................9
5..................................................................................................................................................................................................................10
RTA once crash as dependent variable performing binary logistic regression.....................................................................................10
6..................................................................................................................................................................................................................20
Obesity as dependent variable performing binary logistic regression ..................................................................................................20
2

1.
a. Mean, standard deviation, minimum and maximum values for age
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
AGE 38681 16 59 20.50 4.889
Valid N (listwise) 38681
b. Frequency (% and number) of students in each of new age categories
Statistics
newage
N Valid 32796
Missing 5885
newage
Frequency Percent Valid Percent Cumul
ative
Perce
nt
Valid
.00 11881 30.7 36.2 36.2
1.00 11666 30.2 35.6 71.8
2.00 5494 14.2 16.8 88.6
3.00 3755 9.7 11.4 100.0
Total 32796 84.8 100.0
3

Missing System 5885 15.2
Total 38681 100.0
2.
Descriptive statistics for demographic information
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
cohort 38681 1.00 8.00 5.0695 2.25016
STATE 38681 1.00 4.00 1.8736 .87565
AGE 38681 16 59 20.50 4.889
GENDER 38681 .00 1.00 .7299 .44403
LIVING_ARRANGE 38681 .00 2.00 .7454 .87069
FACULTY 38681 1.00 5.00 2.2809 1.02867
DEGREE_TYPE 38681 .00 1.00 .1050 .30654
METRO 32238 .00 1.00 .1556 .36244
STUDY_MODE 38681 .00 1.00 .1011 .30148
FEE_STATUS 38681 .00 1.00 .1666 .37259
Valid N (listwise) 32238
Interpretation: The above table reflects the descriptive statistics of the demographic information. This is comprised of mean,
minimum, maximum as well as standard deviation. It has been gained that mean of gender is .7299. Along with this its standard
deviation is 0.44. On the other hand the mean value of fee status is 0.1011. Further its minimum value is 0 and maximum value is 1.
However its standard deviation is 0.37
4

3.
Mean for aggression, thrill seeking and risk acceptance differing from various demographics
a. Gender
Group Statistics
GENDER N Mean Std. Deviation Std. Error Mean
driver_agg Male 10449 7.53 2.639 .026
Female 28232 7.54 2.630 .016
thrill Male 10449 3.76 1.879 .018
Female 28232 3.77 1.872 .011
risk_accep Male 10449 8.26 4.488 .044
Female 28232 8.28 4.464 .027
Independen
t Samples
Test
5

Levene's Test for Equality
of Variances
t
-
t
e
s
t
f
o
r
E
q
u
a
l
i
t
y
o
f
M
e
a
n
s
F Sig. t df Sig. (2-
tailed)
Mean
Difference
Std. Error
Difference
95% Confidence Interval of
the Difference
6

Lower Upper
driver_agg
Equal variances
assumed .875 .349 -.494 38679 .621 -.015 .030 -.074 .044
Equal variances
not assumed -.493 18606.923 .622 -.015 .030 -.074 .044
thrill
Equal variances
assumed .897 .344 -.596 38679 .551 -.013 .021 -.055 .029
Equal variances
not assumed -.595 18605.862 .552 -.013 .021 -.055 .029
risk_accep
Equal variances
assumed .922 .337 -.302 38679 .763 -.015 .051 -.116 .085
Equal variances
not assumed -.301 18577.641 .763 -.015 .051 -.116 .085
H0: There is no significant means difference between aggression, thrill seeking and risk acceptance as well as gender
H1: There is significant means difference between aggression, thrill seeking and risk acceptance as well as gender
Interpretation: The table above reflects the determines the difference among the mean for behavior and various demographic factors.
In this regard independent sample T test is being used that allows in determining whether the two sample means are different from one
another on significant basis. From the above table it can be interpreted that in all the cases the significance value is above 0.05. This
implies that H0 is accepted. This implies that there is no significant mean difference between aggression, thrill seeking and risk
acceptance as well as gender.
b. Metropolitan background status
7

Group Statistics
METRO N Mean Std. Deviation Std. Error Mean
driver_agg Metro 27223 7.54 2.629 .016
Non-metro 5015 7.55 2.639 .037
thrill Metro 27223 3.77 1.871 .011
Non-metro 5015 3.79 1.880 .027
risk_accep Metro 27223 8.24 4.477 .027
Non-metro 5015 8.42 4.450 .063
Independen
t Samples
Test
Levene's
Test for
Equality of
Variances
t-test for
Equality of
Means
F S
i
g
.
t d
f
Sig. (2-tailed) Mean
Difference
Std. Error
Difference
9
5
%
C
o
n
f
i
8

d
e
n
c
e
I
n
t
e
r
v
a
l
o
f
t
h
e
D
i
f
f
e
r
e
9

n
c
e
Lower U
p
p
e
r
driver_agg
Equal variances
assumed .164 .68
5 -.174 32236 .862 -.007 .040 -.086 .072
Equal variances not
assumed -.174 6972.542 .862 -.007 .041 -.086 .072
thrill
Equal variances
assumed .330 .56
6 -.529 32236 .597 -.015 .029 -.072 .041
Equal variances not
assumed -.527 6967.182 .598 -.015 .029 -.072 .041
risk_accep
Equal variances
assumed .164 .68
5 -2.487 32236 .013 -.171 .069 -.306 -.036
Equal variances not
assumed -2.497 7012.727 .013 -.171 .068 -.305 -.037
H0: There is no significant means difference between aggression, thrill seeking and risk acceptance as well as Metropolitan
background status
H1: There is significant means difference between aggression, thrill seeking and risk acceptance as well as Metropolitan background
status
10

Interpretation: From the above table it can be interpreted that in cases of aggression and thrill seeking significance value is above
0.05. This implies that H0 is accepted. This implies that there is no significant mean difference between aggression, thrill seeking and
risk acceptance as well as Metropolitan background status. However in case of of risk acceptance the value of significant is 0.013 that
is less than 0.05. Thus in this case alternative hypothesis will be selected. This reflects that there is presence of significant means
difference between aggression, thrill seeking and risk acceptance as well as Metropolitan background status.
c. Study mode
Group Statistics
STUDY_MODE N Mean Std. Deviation Std. Error Mean
driver_agg FT 34770 7.54 2.632 .014
PT 3911 7.55 2.635 .042
thrill FT 34770 3.77 1.873 .010
PT 3911 3.78 1.880 .030
risk_accep FT 34770 8.29 4.475 .024
PT 3911 8.19 4.433 .071
Independen
t Samples
Test
11

Levene's Test for Equality
of Variances
t
-
t
e
s
t
f
o
r
E
q
u
a
l
i
t
y
o
f
M
e
a
n
s
F S
i
t d
f
Sig. (2-tailed) Mean
Difference
Std. Error
Difference
9
5
12

g
.
%
C
o
n
f
i
d
e
n
c
e
I
n
t
e
r
v
a
l
o
f
t
h
e
13

D
i
f
f
e
r
e
n
c
e
Lower U
p
p
e
r
driver_agg
Equal variances
assumed .259 .610 -.34
6 38679 .729 -.015 .044 -.102 .072
Equal variances not
assumed
-.34
6 4830.129 .730 -.015 .044 -.102 .072
thrill
Equal variances
assumed .192 .661 -.17
7 38679 .859 -.006 .032 -.068 .056
Equal variances not
assumed
-.17
7 4825.629 .860 -.006 .032 -.068 .057
risk_accep
Equal variances
assumed
1.08
5 .298 1.24
2 38679 .214 .094 .075 -.054 .241
Equal variances not
assumed
1.25
1 4850.588 .211 .094 .075 -.053 .240
14

H0: There is no significant means difference between aggression, thrill seeking and risk acceptance as well as study mode
H1: There is significant means difference between aggression, thrill seeking and risk acceptance as well as study mode
Interpretation: From the above table it can be interpreted that in all the cases the significance value is above 0.05. This implies that H0
is accepted. This implies that there is no significant mean difference between aggression, thrill seeking and risk acceptance as well as
study mode.
d. RTA in past 12 months
Group Statistics
RTA_one_crash N Mean Std. Deviation Std. Error Mean
driver_agg No RTAs 33931 7.16 2.525 .014
One RTA or more 4750 10.27 1.549 .022
thrill No RTAs 33931 3.50 1.793 .010
One RTA or more 4750 5.72 1.131 .016
risk_accep No RTAs 33931 7.62 4.283 .023
One RTA or more 4750 12.97 2.569 .037
Independen
t Samples
Test
15

Levene's Test for Equality
of Variances
t
-
t
e
s
t
f
o
r
E
q
u
a
l
i
t
y
o
f
M
e
a
n
s
F S
i
t d
f
Sig. (2-tailed) Mean
Difference
Std. Error
Difference
9
5
16

g
.
%
C
o
n
f
i
d
e
n
c
e
I
n
t
e
r
v
a
l
o
f
t
h
e
17

D
i
f
f
e
r
e
n
c
e
Lower U
p
p
e
r
driver_agg
Equal variances
assumed
2246
.668 .000
-
82.8
00
38679 .000 -3.112 .038 -3.186 -3.038
Equal variances not
assumed
-
118.
229
8769.508 .000 -3.112 .026 -3.163 -3.060
thrill
Equal variances
assumed
2424
.557 .000
-
83.2
90
38679 .000 -2.226 .027 -2.279 -2.174
Equal variances not
assumed
-
116.
687
8530.696 .000 -2.226 .019 -2.264 -2.189
18

risk_accep
Equal variances
assumed
2580
.265 .000
-
83.9
86
38679 .000 -5.350 .064 -5.474 -5.225
Equal variances not
assumed
-
121.
767
8974.528 .000 -5.350 .044 -5.436 -5.263
H0: There is no significant means difference between aggression, thrill seeking and risk acceptance as well as RTA in past 12 months
H1: There is significant means difference between aggression, thrill seeking and risk acceptance as well as RTA in past 12 months
Interpretation: From the above table it can be interpreted that in all the cases the significance value is below 0.05. This implies that
H1 is accepted. This implies that there is significant mean difference between aggression, thrill seeking and risk acceptance as well as
RTA in past 12 months.
4.
In terms of depression is there difference among depressed and not depressed group in accordance with following
Group Statistics
depression N Mean Std. Deviation Std. Error Mean
GENDER Not depressed 34958 .7305 .44372 .00237
Depressed 3723 .7241 .44700 .00733
METRO Not depressed 29142 .1552 .36214 .00212
Depressed 3096 .1586 .36535 .00657
STUDY_MODE Not depressed 34958 .1019 .30255 .00162
Depressed 3723 .0935 .29113 .00477
19

FEE_STATUS Not depressed 34958 .1664 .37242 .00199
Depressed 3723 .1684 .37428 .00613
Independen
t Samples
Test
20

Levene's Test for Equality
of Variances
t
-
t
e
s
t
f
o
r
E
q
u
a
l
i
t
y
o
f
M
e
a
n
s
F Sig. t df Sig. (2-
tailed)
Mean
Difference
Std. Error
Difference
9
5
21

%
C
o
n
f
i
d
e
n
c
e
I
n
t
e
r
v
a
l
o
f
t
h
e
22

D
i
f
f
e
r
e
n
c
e
Lower U
p
p
e
r
GENDER
Equal variances
assumed 2.674 .102 .827 38679 .408 .00633 .00766 -.00867 .02133
Equal variances not
assumed .822 4538.839 .411 .00633 .00770 -.00877 .02143
METRO
Equal variances
assumed .950 .330 -.489 32236 .625 -.00335 .00685 -.01678 .01008
Equal variances not
assumed -.486 3770.451 .627 -.00335 .00690 -.01688 .01018
STUDY_MODE
Equal variances
assumed 10.756 .001 1.626 38679 .104 .00845 .00520 -.00174 .01864
Equal variances not
assumed 1.677 4620.908 .094 .00845 .00504 -.00143 .01833
23

FEE_STATUS
Equal variances
assumed .402 .526 -.318 38679 .751 -.00204 .00642 -.01463 .01055
Equal variances not
assumed -.317 4542.907 .752 -.00204 .00645 -.01469 .01060
H0: There is no significant means difference between depression as well as gender, Metropolitan background status, study mode and
fees status
H1: There is significant means difference between depression as well as gender, Metropolitan background status, study mode and fees
status
Interpretation: From the above table it can be interpreted that in all the cases the significance value is above 0.05. This implies that
H0 is accepted. This implies that there is no significant mean difference between depression as well as gender, Metropolitan
background status, study mode and fees status.
5.
RTA one crash as dependent variable performing binary logistic regression
Demographic
Case Processing Summary
Unweighted Casesa N Percent
Selected Cases
Included in Analysis 32796 84.8
Missing Cases 5885 15.2
Total 38681 100.0
Unselected Cases 0 .0
24

Total 38681 100.0
a. If weight is in effect, see
classification table for the total number
of cases.
Dependent Variable Encoding
Original Value Internal Value
No RTAs 0
One RTA or more 1
Categorical Variables
Codings
Frequency Parameter coding
(1) (2)
LIVING_ARRANGE
At home 17355 1.000 .000
College/student accom 5673 .000 1.000
Independently 9768 .000 .000
Classification Tablea,b
Observed Predicted
RTA_one_crash Percentage Correct
No RTAs One RTA or more
Step 0 RTA_one_crash No RTAs 28580 0 100.0
25

One RTA or more 4216 0 .0
Overall Percentage 87.1
a. Constant is included in
the model.
b. The cut value is .500
Variables in the
Equation
B S.E. Wald df Sig. Exp(B)
Step 0 Constant -1.914 .016 13456.877 1 .000 .148
Variables not in the
Equation
Score df Sig.
Step 0 Variables
agecategory 159.500 1 .000
GENDER 54.813 1 .000
LIVING_ARRANGE 16.857 2 .000
LIVING_ARRANGE(1) 5.357 1 .021
LIVING_ARRANGE(2) 16.713 1 .000
FEE_STATUS 41.363 1 .000
Overall Statistics 260.232 5 .000
26

Omnibus Tests of Model
Coefficients
Chi-square df Sig.
Step 1
Step 262.482 5 .000
Block 262.482 5 .000
Model 262.482 5 .000
Model Summary
Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square
1 24900.270a .008 .015
a. Estimation terminated at iteration
number 5 because parameter
estimates changed by less than .001.
Classification Tablea
Observed Predicted
RTA_one_crash Percentage Correct
No RTAs One RTA or more
Step 1 RTA_one_crash No RTAs 28580 0 100.0
One RTA or more 4216 0 .0
Overall Percentage 87.1
a. The cut value is .500
27

Variables in the
Equation
B S.E. Wald df Sig. Exp(B)
Step 1a
agecategory -.236 .018 163.947 1 .000 .789
GENDER -.257 .036 51.121 1 .000 .774
LIVING_ARRANGE 7.516 2 .023
LIVING_ARRANGE(1) -.114 .042 7.252 1 .007 .893
LIVING_ARRANGE(2) -.076 .053 2.053 1 .152 .927
FEE_STATUS .227 .052 18.950 1 .000 1.255
Constant -1.473 .048 925.970 1 .000 .229
a. Variable(s)
entered on step 1:
agecategory,
GENDER,
LIVING_ARRANG
E, FEE_STATUS.
a. In the table above Exp(B) reflects the odds ratio.
b. From the value of living arrangement (1) road traffic accident can be predicted. However living arrangement (2) does not assist in
making prediction of road traffic accident.
It has been determined from Exp (b) that there is presence of direct relationship among the variables that is age gender, living
arrangement and fees status on RTA in the past 12 months
Driving
Case Processing Summary
28

Unweighted Casesa N Percent
Selected Cases
Included in Analysis 38681 100.0
Missing Cases 0 .0
Total 38681 100.0
Unselected Cases 0 .0
Total 38681 100.0
a. If weight is in effect, see
classification table for the total number
of cases.
Dependent Variable Encoding
Original Value Internal Value
No RTAs 0
One RTA or more 1
Classification Tablea,b
29

Observed P
r
e
d
i
c
t
e
d
RTA_one_crash P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
30

No RTAs One RTA or more
Step 0
RTA_one_crash No RTAs 33931 0 100.0
One RTA or more 4750 0 .0
Overall Percentage
8
7
.
7
a. Constant is included in
the model.
b. The cut value is .500
Variables in the
Equation
B S.E. Wald df Sig. E
x
p
(
B
)
Step 0 Constant -1.966 .015 16107.980 1 .000 .140
Variables not in the
Equation
31

Score df S
i
g
.
Step 0
Variables dist_driving 1.704 1 .192
Overall Statistics 1.704 1
.
1
9
2
Omnibus Tests of Model
Coefficients
Chi-square df S
i
g
.
Step 1
Step 1.704 1 .192
Block 1.704 1 .192
Model 1.704 1 .192
Model Summary
Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square
1 28812.976a .000 .000
32

a. Estimation terminated at iteration
number 5 because parameter
estimates changed by less than .001.
Classification Tablea
Observed P
r
e
d
i
c
t
e
d
33

RTA_one_crash P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
No RTAs One RTA or more
Step 1
RTA_one_crash No RTAs 33931 0 100.0
One RTA or more 4750 0 .0
Overall Percentage
8
7
.
7
a. The cut value is .500
34

Variables in the
Equation
B S.E. Wald df Sig. E
x
p
(
B
)
Step 1a dist_driving .040 .031 1.704 1 .192 1.041
Constant -1.987 .022 8082.897 1 .000 .137
a. Variable(s)
entered on step 1:
dist_driving.
It has been determined from Exp (b) that there is presence of direct relationship among the variables that is distance driving on
RTA in the past 12 months. This implies that increase in one will result in increasing another factor to a greater extent.
Behaviour
Case Processing Summary
35

Unweighted Casesa N P
e
r
c
e
n
t
Selected Cases
Included in Analysis 38681 100.0
Missing Cases 0 .0
Total 38681 100.0
Unselected Cases 0 .
0
Total 38681
1
0
0
.
0
a. If weight is in effect, see
classification table for the total number
of cases.
Dependent Variable Encoding
Original Value Internal Value
No RTAs 0
One RTA or more 1
36

Classification Tablea,b
Observed P
r
e
d
i
c
t
e
d
37

RTA_one_crash P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
No RTAs One RTA or more
Step 0
RTA_one_crash No RTAs 33931 0 100.0
One RTA or more 4750 0 .0
Overall Percentage
8
7
.
7
a. Constant is included in
the model.
38

b. The cut value is .500
Variables in the
Equation
B S.E. Wald df Sig. E
x
p
(
B
)
Step 0 Constant -1.966 .015 16107.980 1 .000 .140
Variables not in the
Equation
Score df S
i
g
.
Step 0 Variables driver_agg 5823.908 1 .000
thrill 5882.563 1 .000
risk_accep 5966.062 1 .000
39

Overall Statistics 11865.141 3
.
0
0
0
Omnibus Tests of Model
Coefficients
Chi-square df S
i
g
.
Step 1
Step 16342.624 3 .000
Block 16342.624 3 .000
Model 16342.624 3 .000
Model Summary
Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square
1 12472.057a .345 .656
a. Estimation terminated at iteration
number 8 because parameter
estimates changed by less than .001.
Classification Tablea
40

Observed P
r
e
d
i
c
t
e
d
RTA_one_crash P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
41

No RTAs One RTA or more
Step 1
RTA_one_crash No RTAs 32875 1056 96.9
One RTA or more 1736 3014 63.5
Overall Percentage
9
2
.
8
a. The cut value is .500
Variables in the
Equation
B S.E. Wald df Sig. E
x
p
(
B
)
Step 1a
driver_agg .546 .048 129.792 1 .000 1.727
thrill .650 .066 96.522 1 .000 1.915
risk_accep .601 .010 3995.122 1 .000 1.825
Constant -16.570 .246 4534.099 1 .000 .000
a. Variable(s)
entered on step 1:
driver_agg, thrill,
risk_accep.
42

It has been determined from Exp (b) that there is presence of direct relationship among the variables that is aggression, thrill
seeking as well as risk acceptance on RTA in the past 12 months. This implies that increase in one will result in increasing another
factor to a greater extent.
6.
Obesity as dependent variable performing binary logistic regression
Demographic
Case Processing Summary
Unweighted Casesa N P
e
r
c
e
n
t
Selected Cases
Included in Analysis 32796 84.8
Missing Cases 5885 15.2
Total 38681 100.0
Unselected Cases 0 .
0
Total 38681
1
0
0
.
0
43

a. If weight is in effect, see
classification table for the total number
of cases.
Dependent Variable Encoding
Original Value Internal Value
Not obese 0
Obese 1
Categorical Variables
Codings
44

Frequency P
a
r
a
m
e
t
e
r
c
o
d
i
n
g
(1) (
2
)
LIVING_ARRANGE
At home 17355 1.000 .000
College/student accom 5673 .000 1.000
Independently 9768 .000 .000
Classification Tablea,b
Observed Predict
ed
45

OB P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
Not obese Obese
Step 0 OB Not obese 28580 0 100.0
Obese 4216 0 .0
Overall Percentage 87.1
a. Constant is included in
the model.
b. The cut value is .500
46

Variables in the
Equation
B S.E. Wald df Sig. Exp(B)
Step 0 Constant -1.914 .016 13456.877 1 .000 .148
Variables not in the
Equation
Score df Sig.
Step 0 Variables
agecategory 159.500 1 .000
GENDER 54.813 1 .000
LIVING_ARRANGE 16.857 2 .000
LIVING_ARRANGE(1) 5.357 1 .021
LIVING_ARRANGE(2) 16.713 1 .000
Overall Statistics 240.399 4 .000
Omnibus Tests of Model
Coefficients
Chi-square df Sig.
Step 1
Step 243.688 4 .000
Block 243.688 4 .000
Model 243.688 4 .000
Model Summary
47

Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square
1 24919.064a .007 .014
a. Estimation terminated at iteration
number 5 because parameter
estimates changed by less than .001.
Classification Tablea
Observed Predict
ed
48

OB P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
Not obese Obese
Step 1 OB Not obese 28580 0 100.0
Obese 4216 0 .0
Overall Percentage 87.1
a. The cut value is .500
Variables in the
Equation
49

B S.E. Wald df Sig. E
x
p
(
B
)
Step 1a
agecategory -.240 .018 168.997 1 .000 .787
GENDER -.259 .036 52.215 1 .000 .772
LIVING_ARRANGE 26.054 2 .000
LIVING_ARRANGE(1) -.171 .040 18.338 1 .000 .843
LIVING_ARRANGE(2) .004 .050 .006 1 .936 1.004
Constant -1.412 .046 937.726 1 .000 .244
a. Variable(s)
entered on step 1:
agecategory,
GENDER,
LIVING_ARRANG
E.
b. The significance value reflects that living arrangement (1) is effective in making prediction of obesity. Further the Exp (B) value of
living arrangement demonstrates that it possess direct relationship with obesity. This implies that increase in one will affect another.
Baseline characteristics
Case Processing Summary
50

Unweighted Casesa N P
e
r
c
e
n
t
Selected Cases
Included in Analysis 38681 100.0
Missing Cases 0 .0
Total 38681 100.0
Unselected Cases 0 .
0
Total 38681
1
0
0
.
0
a. If weight is in effect, see
classification table for the total number
of cases.
Dependent Variable Encoding
Original Value Internal Value
Not obese 0
Obese 1
51

Classification Tablea,b
Observed Predict
ed
OB P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
Not obese Obese
Step 0 OB Not obese 33931 0 100.0
Obese 4750 0 .0
52

Overall Percentage 87.7
a. Constant is included in
the model.
b. The cut value is .500
Variables in the
Equation
B S.E. Wald df Sig. E
x
p
(
B
)
Step 0 Constant -1.966 .015 16107.980 1 .000 .140
Variables not in the
Equation
Score df S
i
g
.
Step 0 Variables depression 3503.248 1 .000
53

BL_owob 1.704 1 .192
Overall Statistics 3505.880 2
.
0
0
0
Omnibus Tests of Model
Coefficients
Chi-square df S
i
g
.
Step 1
Step 2495.820 2 .000
Block 2495.820 2 .000
Model 2495.820 2 .000
Model Summary
Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square
1 26318.860a .062 .119
a. Estimation terminated at iteration
number 5 because parameter
estimates changed by less than .001.
54

Classification Tablea
Observed Predict
ed
OB P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
Not obese Obese
Step 1 OB Not obese 33931 0 100.0
Obese 4750 0 .0
55

Overall Percentage 87.7
a. The cut value is .500
Variables in the
Equation
B S.E. Wald df Sig. Exp(B)
Step 1a
depression 2.007 .038 2784.603 1 .000 7.442
BL_owob .055 .033 2.893 1 .089 1.057
Constant -2.335 .025 8717.674 1 .000 .097
a. Variable(s)
entered on step 1:
depression,
BL_owob.
It has been determined from Exp (b) that there is presence of direct relationship among the variables that is overweight,
depression at baseline on obesity. This implies that increase in one will result in increasing another factor to a greater extent.
Parental factors
Case Processing Summary
56

Unweighted Casesa N P
e
r
c
e
n
t
Selected Cases
Included in Analysis 38681 100.0
Missing Cases 0 .0
Total 38681 100.0
Unselected Cases 0 .
0
Total 38681
1
0
0
.
0
a. If weight is in effect, see
classification table for the total number
of cases.
Dependent Variable Encoding
Original Value Internal Value
Not obese 0
Obese 1
57

Classification Tablea,b
Observed Predict
ed
OB P
e
r
c
e
n
t
a
g
e
C
o
r
r
e
c
t
Not obese Obese
Step 0 OB Not obese 33931 0 100.0
Obese 4750 0 .0
58

Overall Percentage 87.7
a. Constant is included in
the model.
b. The cut value is .500
Variables in the
Equation
B S.E. Wald df Sig. E
x
p
(
B
)
Step 0 Constant -1.966 .015 16107.980 1 .000 .140
Variables not in the
Equation
Score df S
i
g
.
Step 0 Variables edu_par 4481.799 1 .000
owob_par 5174.930 1 .000
59

Overall Statistics 5192.598 2
.
0
0
0
Omnibus Tests of Model
Coefficients
Chi-square df S
i
g
.
Step 1
Step 5359.581 2 .000
Block 5359.581 2 .000
Model 5359.581 2 .000
Model Summary
Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square
1 23455.099a .129 .246
a. Estimation terminated at iteration
number 7 because parameter
estimates changed by less than .001.
Classification Tablea
60