Project I

1
Project I

• Fall 2009

2
(No Transcript)
3
(No Transcript)
4
(No Transcript)
5
Bladder
Kidney
Leukemia
Lung
6
(No Transcript)
7
Question 1 Excel, Tools, data Analysis,
Correlation
R2 r2 for bivariate case F1, n-2 R2/1
1-R2 / (n-2) 0.078442/0.9276 F1, 42 3.57,
critical value _at_ 5 F1, 40 4.08
8
Eviews
Gen eff _at_rfdist(1,42) Gen density_at_dfdist(eff,
1, 42)
9
Lung Kidney Correlation
Accept H0
4.08 critical
3.57
10
(No Transcript)
11
Question 1 Excel, Tools, data Analysis,
Correlation
R2 r2 for bivariate case F1, n-2 R2/1
1-R2 / (n-2) 0.078442/0.9276 F1, 42 3.57,
critical value _at_ 5 F1, 40 4.08
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
b.
Cigarettes smoked per Capita is significant,
Prob. 7/10,000 happen by chance
Regression is significant Eventhough R2
0.50, Other unspecified factors Ar work
a.
Coefficent 0.00445, if cigarettes smoked goes up
by one per capita, death rate per 100,000 goes up
by 0.004453
18
If cigarettes smoked per capita goes up by about
1, i.e. by 25, then death rates for lung cancer
go up by 250.0044 0.11125 per 100, 000 or
deaths per year increase by 11,125 Or 0.11125/
19.653, or about 0.6. Called calculating
elasticities at means
19
Or run a log log regression to estimate
elasticity of 0.7
20
4. c
21
(No Transcript)
22
(No Transcript)
23
Copy fitted
24
Paste into a open group window After selecting
edit
25
4. d
2.810-12
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
White Test for Heteroskedasticity
30
In Excel
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
4e.
37
(No Transcript)
38
Q 4e Do DC and Nevada Bias Results?
39
(No Transcript)
40
(No Transcript)
41
b
a
42
5.C Are Residuals Normal?
43
5.d
44
Bladder Cancer and Cigarettes
45
6
46
7
b
a
47
7. C Are the Residuals Normal?
48
Economic and Health Significance

• Smoking is hazardous to your health, especially
  for lung, bladder and Kidney cancer
• Intensity of smoking, e.g cigarettes smoked per
  dollar of income and income per capita affect
  lung cancer death rates
• Smoking intensity may be reduced by advertising
  health hazards and by imposing an excise tax
• Development, as measured by income per capita may
  create dangerous toxins that cause lung cancer,
  e.g. acid rain from coal smoke stacks,
  particlates in the air, smog etc.

49
Beyond The Project
50
Lung a(cigs/pop)/(Inc/pop)Inc/popb
exp(e) lnLung lna b cigs/ b(Inc/pop)
e lnLung lna b(cigs/pop) b(inc/pop)
b(inc/pop) e lnLung lna b(cigs/pop) e
51
(No Transcript)
52
eK
Conjecture for 1960
Kidney cancer Rate
Bladder Cancer Rate
Cigarettes/pop
Income/pop
e
eB
Lung Cancer Rate
eL