14 Let’s return to the carbonation depth-strength data of 13 and calculate a
Example 12.14 Let’s return to the carbonation depth-strength data of Example 12.13 and calculate a
95 PI for a strength value that would result from selecting a single core specimen whose depth is 45 mm. Relevant quantities from that example are
yˆ 5 13.79 s Yˆ 5 .7582 s 5 2.8640
For a prediction level of 95 based on n 2 2 5 16
df, the t critical value is 2.120,
exactly what we previously used for a 95 confidence level. The prediction interval is then
Plausible values for a single observation on strength when depth is 45 mm are (at the
95 prediction level) between 7.51 MPa and 20.07 MPa. The 95 confidence inter- val for mean strength when depth is 45 was (12.18, 15.40). The prediction interval
is much wider than this because of the extra (2.8640) 2 under the square root. Figure
12.18, the Minitab output in Example 12.13, shows this interval as well as the con- fidence interval.
■ The Bonferroni technique can be employed as in the case of confidence inter-
vals. If a 100(1 2 a) PI is calculated for each of k different values of x, the simul- taneous or joint prediction level for all k intervals is at least 100(1 2 ka) .
EXERCISES Section 12.4 (44–56)
44. Fitting the simple linear regression model to the n 5 27
a. Explain why s Yˆ is larger when x 5 60 than when x 5 40 .
observations on x5 modulus of elasticity and y5 flexural
b. Calculate a confidence interval with a confidence level of
strength given in Exercise 15 of Section 12.2 resulted in yˆ 5
95 for the true average strength of all beams whose
7.592, s Yˆ
5 .179 when and x 5 40 yˆ 5 9.741, s Yˆ
modulus of elasticity is 40.
for . x 5 60
CHAPTER 12 Simple Linear Regression and Correlation
c. Calculate a prediction interval with a prediction level of
astringency that can be attributed to the model relation-
95 for the strength of a single beam whose modulus of
ship between astringency and tannin concentration.
elasticity is 40.
b. Calculate and interpret a confidence interval for the slope
d. If a 95 CI is calculated for true average strength when
of the true regression line.
modulus of elasticity is 60, what will be the simultane-
c. Estimate true average astringency when tannin concen-
ous confidence level for both this interval and the inter-
tration is .6, and do so in a way that conveys information
val calculated in part (b)?
about reliability and precision.
45. Reconsider the filtration rate–moisture content data intro-
d. Predict astringency for a single wine sample whose tan-
duced in Example 12.6 (see also Example 12.7).
nin concentration is .6, and do so in a way that conveys
a. Compute a 90 CI for b 0 1 125b 1 , true average mois-
information about reliability and precision.
ture content when the filtration rate is 125.
e. Does it appear that true average astringency for a tannin
b. Predict the value of moisture content for a single experi-
concentration of .7 is something other than 0? State and
mental run in which the filtration rate is 125 using a 90
test the appropriate hypotheses.
prediction level. How does this interval compare to the
47. The simple linear regression model provides a very good fit
interval of part (a)? Why is this the case?
to the data on rainfall and runoff volume given in Exercise
c. How would the intervals of parts (a) and (b) compare to
16 of Section 12.2. The equation of the least squares line is
a CI and PI when filtration rate is 115? Answer without
y 5 21.128 1 .82697x, r 2 5 .975 , and s 5 5.24 .
actually calculating these new intervals.
a. Use the fact that s Yˆ
5 1.44 when rainfall volume is
d. Interpret the hypotheses H 0 :b 0 1 125b 1 5 80 and
40 m 3 to predict runoff in a way that conveys information
H a :b 0 1 125b 1 , 80 , and then carry out a test at sig-
about reliability and precision. Does the resulting inter-
nificance level .01.
val suggest that precise information about the value of
46. Astringency is the quality in a wine that makes the wine
runoff for this future observation is available? Explain
drinker’s mouth feel slightly rough, dry, and puckery. The
your reasoning.
paper “Analysis of Tannins in Red Wine Using Multiple
b. Calculate a PI for runoff when rainfall is 50 using the
Methods: Correlation with Perceived Astringency” (Amer. J.
same prediction level as in part (a). What can be said
of Enol. and Vitic., 2006: 481–485) reported on an investi-
about the simultaneous prediction level for the two in-
gation to assess the relationship between perceived astrin-
tervals you have calculated?
gency and tannin concentration using various analytic
48. The catch basin in a storm-sewer system is the interface
methods. Here is data provided by the authors on x tan-
between surface runoff and the sewer. The catch-basin insert
nin concentration by protein precipitation and y perceived
is a device for retrofitting catch basins to improve pollutant-
astringency as determined by a panel of tasters.
removal properties. The article “An Evaluation of the Urban Stormwater Pollutant Removal Efficiency of Catch Basin
Inserts” (Water Envir. Res., 2005: 500–510) reported on
tests of various inserts under controlled conditions for which inflow is close to what can be expected in the field.
Consider the following data, read from a graph in the arti-
.336 cle, for one particular type of insert on x .221 .898 .836 amount filtered
(1000s of liters) and y total suspended solids removed.
.090 1.132 .538 1.098 .581 .862 .551 Summary quantities are
5 1251, g x i 5 199,365, g y i 5 250.6,
Relevant summary quantities are as follows:
g y 2 i 5 9249.36, g x i y i 5 21,904.4
gx i 5 19.404, g y
i
5 2.549, g x i 5 13.248032,
g y 2
a. Does a scatter plot support the choice of the simple lin-
i
5 11.835795, g x i y i 5 3.497811
ear regression model? Explain.
b. Obtain the equation of the least squares line.
S
5 13.248032 2 (19.404) xx 2 32 5 1.48193150,
c. What proportion of observed variation in removed can
S yy 5 11.82637622
be attributed to the model relationship? d. Does the simple linear regression model specify a useful
S xy 5 3.497811 2 (19.404)(2.549)32
relationship? Carry out an appropriate test of hypotheses
using a significance level of .05.
a. Fit the simple linear regression model to this data. Then
e. Is there strong evidence for concluding that there is at least
determine the proportion of observed variation in
a 2 decrease in true average suspended solid removal
12.4 Inferences Concerning m Y • x and the Prediction of Future Y Values
associated with a 10,000 liter increase in the amount fil-
b. Use the Minitab output from the example to calculate
tered? Test appropriate hypotheses using a 5 .05 .
a 95 CI for expected rolling speed when test
f. Calculate and interpret a 95 CI for true average
speed
removed when amount filtered is 100,000 liters. How
c. Use the Minitab output to calculate a 95 PI for a single
does this interval compare in width to a CI when amount
value of rolling speed when test speed = 47.
filtered is 200,000 liters?
g. Calculate and interpret a 95 PI for removed when
52. Plasma etching is essential to the fine-line pattern transfer in
amount filtered is 100,000 liters. How does this interval
current semiconductor processes. The article “Ion Beam-
compare in width to the CI calculated in (f) and to a PI
Assisted Etching of Aluminum with Chlorine” (J. of the
when amount filtered is 200,000 liters?
Electrochem. Soc., 1985: 2010– 2012) gives the accompa- nying data (read from a graph) on chlorine flow (x, in
49. You are told that a 95 CI for expected lead content when
SCCM) through a nozzle used in the etching mechanism
traffic flow is 15, based on a sample of n 5 10 observa-
and etch rate (y, in 100 Amin).
tions, is (462.1, 597.7). Calculate a CI with confidence level
99 for expected lead content when traffic flow is 15.
x
50. Silicon-germanium alloys have been used in certain types of
y
solar cells. The paper “Silicon-Germanium Films Deposited by Low-Frequency Plasma-Enhanced Chemical Vapor Deposition” (J. of Material Res., 2006: 88–104) reported on a
The summary statistics are gx i 5 24.0, g y i 5 312.5, 2 study of various structural and electrical properties. Consider 2 gx i 5 70.50, g x i y i 5 902.25, g y i 5 11,626.75, bˆ 0 5
the accompanying data on x Ge concentration in solid phase
6.448718, bˆ 1 5 10.602564 .
(ranging from 0 to 1) and y Fermi level position (eV):
a. Does the simple linear regression model specify a useful
x
0 .42 .23 .33 .62 .60 .45 .87 .90 .79 1 relationship between chlorine flow and etch rate? b. Estimate the true average change in etch rate associated
y .62 .53 .61 .59 .50 .55 .59 .31 .43 .46 .23 .22 .19
with a 1-SCCM increase in flow rate using a 95 confi- dence interval, and interpret the interval.
A scatter plot shows a substantial linear relationship. Here
c. Calculate a 95 CI for m Y 3.0 , the true average etch rate
is Minitab output from a least squares fit. [Note: There are
when flow
5 3.0. Has this average been precisely
several inconsistencies between the data given in the paper,
estimated?
the plot that appears there, and the summary information
d. Calculate a 95 PI for a single future observation on
about a regression analysis.]
etch rate to be made when flow 5 3.0 . Is the prediction
The regression equation is
likely to be accurate?
Fermi pos 5 0.7217 2 0.4327 Ge conc
e. Would the 95 CI and PI when flow 5 2.5 be wider or S 5 0.0737573 R – Sq 5 80.2 R – Sq(adj) 5 78.4 narrower than the corresponding intervals of parts
(c) and (d)? Answer without actually computing the intervals.
Analysis of Variance
f. Would you recommend calculating a 95 PI for a flow
of 6.0? Explain.
53. Consider the following four intervals based on the data of
a. Obtain an interval estimate of the expected change in
Exercise 12.17 (Section 12.2):
Fermi-level position associated with an increase of .1 in
a. A 95 CI for mean porosity when unit weight is 110
Ge concentration, and interpret your estimate.
b. A 95 PI for porosity when unit weight is 110
b. Obtain an interval estimate for mean Fermi-level position
c. A 95 CI for mean porosity when unit weight is 115
when concentration is .50, and interpret your estimate.
d. A 95 PI for porosity when unit weight is 115
c. Obtain an interval of plausible values for position result-
Without computing any of these intervals, what can be said
ing from a single observation to be made when concen-
about their widths relative to one another?
tration is .50, interpret your interval, and compare to the
54. The decline of water supplies in certain areas of
interval of (b).
the United States has created the need for increased
d. Obtain simultaneous CIs for expected position when
understanding of relationships between economic factors
concentration is .3, .5, and .7; the joint confidence level
such as crop yield and hydrologic and soil factors. The
should be at least 97.
article “Variability of Soil Water Properties and Crop
51. Refer to Example 12.12 in which x test track speed and y
Yield in a Sloped Watershed” (Water Resources Bull.,
rolling test speed.
1988: 281–288) gives data on grain sorghum yield (y, in
a. Minitab gave s bˆ 0 1bˆ 1 (45) 5 .120 and s bˆ 0 1bˆ 1 (47) 5 .186 . Why
gm-row) and distance upslope (x, in m) on a sloping
is the former estimated standard deviation smaller than
watershed. Selected observations are given in the accom-
the latter one?
panying table.
CHAPTER 12 Simple Linear Regression and Correlation
a. Is it plausible that yield load is normally distributed?
b. Estimate true average yield load by calculating a confi- dence interval with a confidence level of 95, and inter-
pret the interval.
c. Here is output from Minitab for the regression of yield
a. Construct a scatter plot. Does the simple linear regres-
load on torque. Does the simple linear regression model
sion model appear to be plausible?
specify a useful relationship between the variables?
b. Carry out a test of model utility. c. Estimate true average yield when distance upslope is 75
Predictor
Coef
SE Coef
T
P
by giving an interval of plausible values.
55. Verify that V(bˆ 0 1 bˆ 1 x) is indeed given by the expression in
the text. [Hint: V(g d i Y i )5gd 2 i V(Y i ) .]
S 5 73.2141 R – Sq 5 53.6 R – Sq(adj) 5 50.0
56. The article “Bone Density and Insertion Torque as
F Predictors of Anterior Cruciate Ligament Graft Fixation P
Strength” (The Amer. J. of Sports Med., 2004: 1421–1429)
Residual Error 13 69684
gave the accompanying data on maximum insertion torque
Total
(N m) and yield load (N), the latter being one measure of
d. The authors of the cited paper state, “Consequently, we
graft strength, for 15 different specimens.
cannot but conclude that simple regression analysis- based methods are not clinically sufficient to predict Torque 1.8 2.2 1.9 1.3 2.1 2.2 1.6 2.1 individual fixation strength.” Do you agree? [Hint:
Load
Consider predicting yield load when torque is 2.0.]