6. Risk Factor of Stroke
Image Source: http://www.easierliving.com/blog/2013/05/14/know-how-to-prevent-recurrent-stroke/
STROKE
7. Risk Factor of Stroke
Image Source: http://www.easierliving.com/blog/2013/05/14/know-how-to-prevent-recurrent-stroke/
STROKE
8. Risk Factor of Stroke
Image Source: http://www.easierliving.com/blog/2013/05/14/know-how-to-prevent-recurrent-stroke/
STROKE
Controlled
9. Research Question
How do the prevalence of diabetes
influence the prevalence of stroke
in Indonesia’s province, given each
province has constant prevalence
of hypertension ?
12. Where is the study population?
Image Source: http://www.pksbalikpapantengah.org/2013/05/wujudkan-indonesia-sebagai-sepenggal.html
13. How is the data collected?
Data Resource:
Data Resource: http://labmandat.litbang.depkes.go.id/
14. How is the data collected?
Data Resource: http://labmandat.litbang.depkes.go.id/
15. How is the data collected?
Data Resource: http://labmandat.litbang.depkes.go.id/
16. How is the data collected?
Data Resource: http://labmandat.litbang.depkes.go.id/
17. How is the data collected?
Data Resource: http://labmandat.litbang.depkes.go.id/
18. How is the data collected?
Data Resource: http://labmandat.litbang.depkes.go.id/
Prevalence of
Stroke
19. How is the data collected?
Data Resource: http://labmandat.litbang.depkes.go.id/
Prevalence of
Diabetes
Prevalence of
Stroke
20. How is the data collected?
Data Resource: http://labmandat.litbang.depkes.go.id/
Prevalence of
Diabetes
Prevalence of
Stroke
Prevalence of
Hypertension
21. How is the data analyzed?
Statistical Tool & Model
22. How is the data analyzed?
Y : prevalence of stroke
X : prevalence of diabetes
X : prevalence of hypertension
i
1i
2i
Statistical Tool & Model
23. How is the data analyzed?
Y : prevalence of stroke
X : prevalence of diabetes
X : prevalence of hypertension
i
1i
2i
Statistical Tool & Model
E(Y |X , X ) = βX + βX + α (constant) + ɛ (residual error)1i 2i 1i 2ii
24. How is the data analyzed?
Y : prevalence of stroke
X : prevalence of diabetes
X : prevalence of hypertension
i
1i
2i
Statistical Tool & Model
STATA ®12
E(Y |X , X ) = βX + βX + α (constant) + ɛ (residual error)1i 2i 1i 2ii
30. Regression Assumption
1) Independent observation
2) Linear relationship
3) Homoscedasticity (constant variance)
4) Y | X, is normaly distributed
31. 5
101520
.5 1 1.5 2 2.5
Prevalence of Diabetes
Regression fit Prevalence of Stroke
Prevalenceofstroke
Prevalence of diabetes
r = 0.61 (p < 0.001)
Peason’s correlation analysis
32. 5
101520
.5 1 1.5 2 2.5
Prevalence of Diabetes
Regression fit Prevalence of Stroke
Prevalenceofstroke
Prevalence of diabetes
r = 0.61 (p < 0.001)
Multivariate linear regression
E(Y | X ,X ) = 6.11 + 4.45 X - 0.083X + ɛ ~N (0, 2.5 )1i 2i1i 2ii
Prevalence of stroke increased significantly with prevalence of diabetes
(t(30) = 4.25 ,p=0.000).
2
33. 5
101520
.5 1 1.5 2 2.5
Prevalence of Diabetes
Regression fit Prevalence of Stroke
Prevalenceofstroke
Prevalence of diabetes
r = 0.61 (p < 0.001)
Multivariate linear regression
E(Y | X ,X ) = 6.11 + 4.45 X - 0.083X + ɛ ~N (0, 2.5 )1i 2i1i 2ii
Prevalence of stroke increased significantly with prevalence of diabetes
(t(30) = 4.25 ,p=0.000).
2
Adj-R = 0.35
2
34. 5
101520
.5 1 1.5 2 2.5
Prevalence of Diabetes
Regression fit Prevalence of Stroke
Prevalenceofstroke
Prevalence of diabetes
r = 0.61 (p < 0.001)
Multivariate linear regression
E(Y | X ,X ) = 6.11 + 4.45 X - 0.083X + ɛ ~N (0, 2.5 )1i 2i1i 2ii
Prevalence of stroke increased significantly with prevalence of diabetes
(t(30) = 4.25 ,p=0.000).
Y : prevalence of strokei
2
Adj-R = 0.35
2
35. 5
101520
.5 1 1.5 2 2.5
Prevalence of Diabetes
Regression fit Prevalence of Stroke
Prevalenceofstroke
Prevalence of diabetes
r = 0.61 (p < 0.001)
Multivariate linear regression
E(Y | X ,X ) = 6.11 + 4.45 X - 0.083X + ɛ ~N (0, 2.5 )1i 2i1i 2ii
Prevalence of stroke increased significantly with prevalence of diabetes
(t(30) = 4.25 ,p=0.000).
Y : prevalence of strokei
X : prevalence of diabetes1i
2
Adj-R = 0.35
2
36. 5
101520
.5 1 1.5 2 2.5
Prevalence of Diabetes
Regression fit Prevalence of Stroke
Prevalenceofstroke
Prevalence of diabetes
r = 0.61 (p < 0.001)
Multivariate linear regression
E(Y | X ,X ) = 6.11 + 4.45 X - 0.083X + ɛ ~N (0, 2.5 )1i 2i1i 2ii
Prevalence of stroke increased significantly with prevalence of diabetes
(t(30) = 4.25 ,p=0.000).
Y : prevalence of strokei
X : prevalence of diabetes1i
X : prevalence of hypertension2i
2
Adj-R = 0.35
2
37. 5
101520
.5 1 1.5 2 2.5
Prevalence of Diabetes
95% CI Regression fit
Prevalence of Stroke
Prevalenceofstroke
Prevalence of diabetes
r = 0.61 (p < 0.001)
Multivariate linear regression
E(Y | X ,X ) = 6.11 + 4.45 X - 0.083X + ɛ ~N (0, 2.5 )1i 2i1i 2ii
Prevalence of stroke increased significantly with prevalence of diabetes
(t(30) = 4.25 ,p=0.000). The increase was estimated to be 4.45% (95% CI from 2.31% -
6.59%) per 1 % increased of prevalence of diabetes,
2
Adj-R = 0.35
2
38. 5
101520
.5 1 1.5 2 2.5
Prevalence of Diabetes
95% CI Regression fit
Prevalence of Stroke
Prevalenceofstroke
Prevalence of diabetes
r = 0.61 (p < 0.001)
Multivariate linear regression
E(Y | X ,X ) = 6.11 + 4.45 X - 0.083X + ɛ ~N (0, 2.5 )1i 2i1i 2ii
Prevalence of stroke increased significantly with prevalence of diabetes
(t(30) = 4.25 ,p=0.000). The increase was estimated to be 4.45% (95% CI from 2.31% -
6.59%) per 1 % increased of prevalence of diabetes, given controlling prevalence of
hypertension (within province has constant prevalence of hypertension).
2
Adj-R = 0.35
2