In: Statistics and Probability
The relationship between the amount of remaining carbon monoxide (CO) in an individual’s lungs and the time since that person last smoked a cigarette can be summarized using a linear regression approach.
(a) Write down a simple linear regression model and the underlying assumptions. The following summary data was collected from 12 different smokers.
x = time since last smoked a cigarette (hours) y = amount of CO in ppm
n=12, x=1.88 sxx =25.8, syy =1805 SSE=877.4 Fitted regression line: ?̂ = ??. ?? − ?. ??? ?
(b) State the values of ? and ? from the fitted regression equation. Not that ? and ? are the estimates of intercept parameter ? and slope parameter ? respectively.
(c) What is the estimated change in the remaining CO in the lungs if the time since last smoked is increased by 1 hour?
(d) State the hypotheses to test the significance of the linear regression.
(e) Compute MSE.
Page 4 of 5
(f) Calculate the test statistic, state the degrees of freedom, and find the p-value to test the hypotheses in (c).
(g) Would you reject H0 or fail to reject H0 at 5% level of significance?
(h) What do you conclude about the regression of CO in the lungs on time since last
smoked?
(i) Compute the coefficient of determination, r2 using 1-SSE/Syy. Interpret the value.
(j) Predict the amount of remaining CO in the lungs for an elapsed time of 2.25 hours
after smoking.
(k) What is the residual at ?0 = 2.25 if the corresponding observed amount CO is
28ppm?
(l) Estimate the mean amount of CO in the lungs for an elapsed time of 2.25 hours.
(m) Construct a 95% confidence interval for the true mean amount of CO in the lungs
for an elapsed time of 2.25 hours.
Answer:
a) Ŷ = 45.1700 + -5.9960 *x
b)
estimated slope , b = -5.99600
intercept, a = 45.17000
c) decrease by 5.996
d)
Ho: ß1= 0
H1: ß1╪ 0
e) MSE=SSE/df = 877.4/10 = 87.74
f) test stat=10.5721
p value=0.0087
g) reject ho
i) R²= SSR/SST=0.5139
about 51.39% of variation in observation of variably y, is xplained by variable x
j)
Predicted Y at X= 2.25 is
Ŷ = 45.1700 +
-5.9960 *2.25= 31.679
k) residual = 28-31.679 = -3.6790
l) Ŷ = 45.1700 + -5.9960 *2.25= 31.679
m)
X Value= 2.25
Confidence Level= 95%
Sample Size , n= 12
Degrees of Freedom,df=n-2 = 10
critical t Value=tα/2 = 2.228 [excel
function: =t.inv.2t(α/2,df) ]
X̅ = 1.88
Σ(x-x̅)² =Sxx 25.80
Standard Error of the Estimate,Se= 9.3670
Predicted Y at X= 2.25 is
Ŷ = 45.1700 +
-5.9960 *2.25= 31.679
standard error, S(ŷ)=Se*√(1/n+(X-X̅)²/Sxx) =
2.789
margin of error,E=t*Std error=t* S(ŷ) =
2.2281 * 2.789 =
6.2138
Confidence Lower Limit=Ŷ +E =
31.679 - 6.214 =
25.465
Confidence Upper Limit=Ŷ +E = 31.679
+ 6.214 = 37.893
Please upvote if you have liked my answer, would be of great help. Thank you