In: Statistics and Probability
Cardiologists use the short-range scaling exponent α1, which measures the randomness of heart rate patterns, as a tool to assess risk of heart attack. The article “Applying Fractal Analysis to Short Sets of Heart Rate Variability Data” compared values of α1 computed from long series of measurements (approximately 40,000 heartbeats) with those estimated from the first 300 beats to determine how well the long-term measurement (y) could be predicted the short-term one (x). Following are the data (obtained by digitizing a graph).
Short |
Long |
0.5 |
0.6 |
1.02 |
0.79 |
1.4 |
0.81 |
0.88 |
0.9 |
1.68 |
1.05 |
1.16 |
1.05 |
0.82 |
1.05 |
0.93 |
1.07 |
1.26 |
1.1 |
1.18 |
1.19 |
0.81 |
1.19 |
0.81 |
1.2 |
1.28 |
1.23 |
1.18 |
1.23 |
0.71 |
1.24 |
Note: This problem has a reduced data set for ease of performing the calculations required. This differs from the data set given for this problem in the text.
A) Find a 95% confidence interval for the mean long-term measurement for those with short-term measurements of 1.2. Round the answers to three decimal places.
B) Find a 95% prediction interval for the long-term measurement for a particular individual whose short term measurement is 1.2. Round the answers to three decimal places.
For A:
(0.938, 1.188) = Incorrect
(0.936, 1.185) = Incorrect
x | y | (x-xbar)^2 | (y-ybar)^2 | (x-xbar)*(y-ybar) | |
0.5 | 0.6 | 0.2930418 | 0.1995111 | 0.241795556 | |
1.02 | 0.79 | 0.0004551 | 0.0658778 | 0.005475556 | |
1.4 | 0.81 | 0.1286418 | 0.0560111 | -0.084884444 | |
0.88 | 0.9 | 0.0260284 | 0.0215111 | 0.023662222 | |
1.68 | 1.05 | 0.4078951 | 1.111E-05 | 0.002128889 | |
1.16 | 1.05 | 0.0140818 | 1.111E-05 | 0.000395556 | |
0.82 | 1.05 | 0.0489884 | 1.111E-05 | -0.000737778 | |
0.93 | 1.07 | 0.0123951 | 0.0005444 | -0.002597778 | |
1.26 | 1.1 | 0.0478151 | 0.0028444 | 0.011662222 | |
1.18 | 1.19 | 0.0192284 | 0.0205444 | 0.019875556 | |
0.81 | 1.19 | 0.0535151 | 0.0205444 | -0.033157778 | |
0.81 | 1.2 | 0.0535151 | 0.0235111 | -0.035471111 | |
1.28 | 1.23 | 0.0569618 | 0.0336111 | 0.043755556 | |
1.18 | 1.23 | 0.0192284 | 0.0336111 | 0.025422222 | |
0.71 | 1.24 | 0.1097818 | 0.0373778 | -0.064057778 | |
sum | 15.62 | 15.7 | 1.2915733 | 0.5155333 | 0.153266667 |
mean | 1.041333333 | 1.046666667 | sxx | syy | sxy |
slope=b1=sxy/sxx | 0.118666639 | 0.4746666 | 0.665849106 | ||
intercept=b0=ybar-(slope*xbar) | 0.92309514 | 1.3977617 | 0.18782777 | ||
SST | SYY | 0.5155333 | |||
SSR | sxy^2/sxx | 0.0181876 | 0.18782777 | ||
SSE | syy-sxy^2/sxx | 0.4973457 | 0.035279271 | ||
r^2 | SSR/SST | 0.0352793 | |||
error variance s^2 | SSE/(n-2) | 0.0382574 | |||
S^2b1 | s^2/sxx | 0.0296207 | |||
standard error b1=se(b1) | sqrt(s^2b1) | 0.1721068 |
estimated slope , ß1 = SSxy/SSxx = 0.1533
/ 1.2916 = 0.1187
intercept, ß0 = y̅-ß1* x̄ = 0.9231
so, regression line is Ŷ
= 0.9230 + 0.1187
*x
SSE= (SSxx * SSyy - SS²xy)/SSxx =
0.4973
std error ,Se = √(SSE/(n-2)) =
0.1956
========
A)
X Value= 1.2
Confidence Level= 95%
Sample Size , n= 15
Degrees of Freedom,df=n-2 = 13
critical t Value=tα/2 = 2.160 [excel
function: =t.inv.2t(α/2,df) ]
X̅ = 1.04
Σ(x-x̅)² =Sxx 1.29
Standard Error of the Estimate,Se= 0.1956
Predicted Y at X= 1.2 is
Ŷ = 0.9230 +
0.1187 * 1.2 =
1.065
standard error, S(ŷ)=Se*√(1/n+(X-X̅)²/Sxx) = 0.0574
margin of error,E=t*Std error=t* S(ŷ) =
2.1604 * 0.0574 =
0.1240
Confidence Lower Limit=Ŷ +E =
1.065 -
0.1240 = 0.941
Confidence Upper Limit=Ŷ +E =
1.065 + 0.1240
= 1.189
==============
B)
For Individual Response Y
standard error, S(ŷ)=Se*√(1+1/n+(X-X̅)²/Sxx) = 0.20
margin of error,E=t*std error=t*S(ŷ)=
2.1604 * 0.20 = 0.4404
Prediction Interval Lower Limit=Ŷ -E
= 1.065
- 0.44 = 0.625
Prediction Interval Upper Limit=Ŷ +E =
1.065 + 0.44
= 1.505