On a certain country road that runs from a coastal town to a village in the

mountains, the average speed of motor cars is 80 km/hr uphill and 100 km/hr downhill. What is the
average speed for a journey from the town to the village and back?

Let distance = kilometers

Average speed = 2 /( /80+ /100) = 88.9 kilometers per hour, using the harmonic
mean.

Catchment-averaged annual rainfall in the Po River basin of Italy for the 61-
year period from 1918 to 1978 are given in the penultimate column of Table E.7.2. Draw a stem-and-
leaf plot and a box plot of the data. Comment on the type of distribution.

Ranked mean annual rainfall in mm in the Po river basin in Italy from 1918 to 1978:
807 846 876 885 886 896 909 913 922 940
940 950 959 969 978 986 987 993 995 997
999 1011 1015 1017 1026 1028 1029 1046 1046 1051
1090 1096 1100 1110 1112 1123 1133 1133 1142 1159
1171 1196 1197 1215 1210 1228 1259 1264 1290 1318
1323 1345 1349 1356 1362 1422 1496 1501 1529 1564
1654
= 61. Median 2 = 1,090 mm. Medians of the bottom and top halves of the ranked
data:
1 = (978 + 986)/2 = 982 mm; 3 = (1,228 + 1,259)/2 = 1,243.5 mm:

Range = 1,654 – 807 = 847 mm; iqr = 3 - 1 = 1,243.5 – 982 = 261.5 mm

.To demarcate high outliers add 1.5 iqr
= 1.5 ×261.5 = 392.25 mm to 3 ; hence demarcation point ≈ 1,636 mm

Stem-and-leaf plot
5 8⏐1 5 8 9 9
18 9⏐0 1 1 2 4 5 5 6 7 8 9 9 9
(13) 10⏐0 0 0 1 2 2 3 3 3 5 5 5 9
30 11⏐0 0 0 1 3 3 3 4 6 7
20 12⏐0 0 2 2 3 6 6 9
12 13⏐2 2 5 5 6 6
6 14⏐2
5 15⏐0 0 3 6
1 16⏐5

Sketch of box and whiskers plot:

(Data signposts:lowest value, 3 quartiles, high outlier limit, 1 outlier)
_______________
⏐ ⏐ ⏐
⏐ ⏐______⏐______⏐ ⏐ ⏐
807 982 1,090 1,243.5 1,636 1,654 mm
1 2 3 Quartiles
800 900 1,000 1,100 1,200 1,300 1,400 1,500 1,600 1,700mm
Comments: The distribution is not symmetrical. It has a long right tail, The highest
value is a suspected outlier.
 Water quality measurements are taken daily on the River Ouse at Clapham,
England. The concentrations of chlorides and phosphates in solution, given below in milligrams per
liter, are determined over a 30-day period.

semua diganti 6b Chloride: 64.0, 66.0, 64.0, 62.0, 65.0, 64.0, 64.0, 65.0, 65.0, 67.0, 67.0, 74.0 69.0, 68.0, 68.0,
69.0, 63.0, 68.0, 66.0, 66.0, 65.0, 64.0, 63.0, 66.0, 55.0, 69.0, 65.0, 61.0, 62.0, 62.0
Phosphate: 1.31, 1.39, 1.59, 1.68, 1.89, 1.98, 1.97, 1.99, 1.98, 2.15, 2.12, 1.90 1.92, 2.00,
semua diganti...,ab 1.90, 1.74, 1.81, 1.86, 1.86, 1.65, 1.58, 1.74, 1.89, 1.94, 2.07, 1.58, 1.93, 1.72, 1.73, 1.82

Compare the coefficients of variation . Draw a scatter diagram and compute the correlation coefficient
. Comment on the results. Do you see any role in this association for predictive purposes?

Coefficients of variation of 4.9% and 10.7%. Scatter diagram follows. = 0.027.

There is no role in the association for predictive purposes.

 From the timber strength data of Table 1.1.3, compute the 3 percent trimmed
mean by omitting 3 percent of the observations from the highest and the lowest extremities of the
ranked data. Compute the standard deviation $ and the coefficients of skewness 1 and kurtosis 2.
Compare with the results for the full sample (as given in Table 1.2.2).

Mean Std dev Coef. of variation Coef of skew Coef of kurt

Sample 3% trimmed 8.94 7.94 20% 0.14 2.48
Full sample 39.09 9.92 25% 0.15 4.46

It is seen that trimming reduces all the statistics.

Joist-hanger tests carried out at the University of Birmingham, England, on

concrete beams gave observations of deflections in millimeters and failure load in kilograms. The
following results pertain to 75 mm × 150 mm hangers on which timber joists rest:

Failure load: 1903, 1665, 1903, 1991, 2229, 1910, 2025, 1991, 1882, 2032, 1896, 1346
Deflection: 0.69, 0.67, 0.80, 0.50, 0.74, 0.78, 0.57, 0.91, 0.54, 0.50, 0.97, 0.62

Determine by drawing a scatter diagram and computing the correlation coefficients whether there is
any association between the two variables. Discuss your results.

Scatter diagram follows.

= 0.069. There is no role in the association for predictive purposes.

 Applied Statistics for Civil and Environmental Engineers
by N.T. Kottegoda and R. Rosso ©

The following are measurements of concentrations of chloride in

milligrams per liter in a shallow unconfined aquifer taken at intervals of three months [from J. Harris,
J.C. Loftis, and R.H. Montgomery (1987). “Statistical models for characterizing ground-water quality”,
Groundwater, Vol. 25, pp. 185-193]:
38 40 35 37 32 37 37 32 45 38
33.8 14 39 46 48 41 35 49 64 73
67 67 59 73 92.5 45.5 40.4 33.9 28.1
Compute the coefficients of skewness and kurtosis and make an approximate test for normality, using
α = .05. (Data used with the kind courtesy of the publishers).

For groundwater quality, sample size n= 29; sample coefficient of skewness g1 =

0.9660 and sample coefficient of kurtosis = 3.737.
6n (n − 1) 6 × 29 × 28
Var[g1 ] = = = 0.187
(n − 2)( n + 1)(n + 3) 27 × 30 × 32

For the population coefficient of skewness γ 1 , the 95% confidence limits

are±1.96× 0.187 = ± 0.85
For the population coefficient of skewness γ 1 , the approximate 95% confidence limits
are (from the text, for the given sample size) are 4.60, 1.85.
Reject NH of normality in distribution.

. Annual rainfall from 1918 to 1978 in the Po

River basin of northern Italy are given in the penultimate column of Table E.7.2. Divide the record into
two parts of 30 and 31 years. Determine whether the rainfall regime has changed by testing whether the
two parts belong to the same population at the 5 percent level of significance using the Kolmogorov-
Smirnov two-sample test.

Po annual rainfall from Table E7.2

First half of data m = 30:- 1133 999 1501 807 1051 969 997 1090 1356 1133
1171 876 1159 993 1112 1128 1345 1290 1259 1529
940 1196 1046 1218 948 896 950 846 1011 1096
Second half of data n =31:-1100 922 978 1496 913 1046 1100 886 1028 1215
1142 1422 1654 987 909 1362 1026 1015 1228 885
1264 995 986 1017 1349 1029 959 1323 1318 1564 1197
The (critical) d m,n occurs after the 24th item of the ranked data of the first part with m
= 30, and the 19th item of the ranked data of the second part with n = 31.

Problem Solution Manual for Chapter 5 - Page 24 (out of 38)

 Applied Statistics for Civil and Environmental Engineers
by N.T. Kottegoda and R. Rosso ©
1/ 2 1/ 2
⎡ mn ⎤ ⎡ 30 × 31⎤
d m,n = 0.1871 and ⎢⎣ m + n ⎥⎦ × d m,n = ⎢ × 1871 = 0.7305 .
⎣ 61 ⎥⎦
This is low when compared with the critical value in Table C.7. Therefore do not
reject the NH that the two parts belong to the same population.

The following are the ranked annual inflows in 106, for the period 1950 to 1974,
to the Warragamba reservoir, which supplies water to the city of Sydney, Australia:
724 1,505 3,310 6,551 6,915 7,114 7,811 8,962 9,219 9,664
9,840 10,134 10,299 10,824 11,953 12,566 13,969 14,941 15,449 16,800
17,601 18,250 18,483 19,081 20,242
(By kind courtesy of the University of New South Wales, Sydney)
Test whether the distribution is normal using Table C.8 of Appendix C, which is Lilliefors’ test for
normality corrected by Dallal and Wilkinson (1986) for the purpose.

724 10134

k 1 2 3 4 5 6 7 8 9 10 11 12
Fn (x) .04 .08 .12 .16 .20 .24 .28 .32 .36 .40 .44 .48

F0 ( x) .026 .036 .071 .192 .211 .221 .261 .334 .352 .383 .395 .416

10824 11953

k 13 14 15 16 17 18 19 20 21 22 23 24 25
Fn (x) .52 .56 .60 .64 .68 .72 .76 .80 .84 .88 .92 .96 1.00

F0 ( x) .428 .466 .549 .593 .689 .75 .78 .84 .88 .90 .91 .92 .95

The critical value d n = 0.094, which occurs at k = 14. From Table C.8, the limiting
value Dn , 0.05 =0.173. Therefore do not reject the NH that the distribution is normal

Problem Solution Manual for Chapter 5 - Page 25 (out of 38)

 Applied Statistics for Civil and Environmental Engineers
BY N.T. KOTTEGODA AND R. ROSSO ©

The straight line

P = 11.145Y + 82.8 provides a good approximation
t 8,0.025 = 2.306.
Confidence limits for y =15:-
1/ 2
⎡ 1 (15 − 5.5) 2 ⎤
11.145× 15 + 82.8 ± 2.306 15.5819 ⎢1 + + ⎥ ⇒ 240,260
⎣ 10 82.5 ⎦

For the water quality measurements on the River Ouse at Clapham, England, data
given in Problem 1.15 (Chapter 1) determine the linear regression equation by least squares using
phosphate as the explanatory variable. Comment on the model and the results.

The linear regression model (see Eqs. 6.1 to 6.7) takes the form
CHL =0.454 phos +64.4 + ε
With r = 0.03, this serves no purpose for prediction purposes.

The lowest annual flows measured, in cubic meters per second, at stations
X and Y on the Jackson and Cowpasture rivers, respectively, in the United States are to be correlated in
order to extend the shorter record at station Y. A simple linear regression model is to be used. The
following summary statistics have been computed over a 12-year period:
Sum X = 28.77; Sum Y = 28.23; Sum XX = 73.14; Sum YY = 71.20; Sum XY = 71.53.
(1) Find least squares estimates of the parameters.
(2) What is the standard error of the residuals?
(3) Estimate the coefficient of correlation.
(4) Find approximate 95 percent confidence limits for the population correlation coefficient.

S xy Σxy − (ΣxΣy ) / n 71.53 − 28.77 × 28.23 / 12

(a) βˆ1 = = = = 0.924 .
S xx Σ x 2 − (Σx) 2 / n 73.14 − 28.77 2 / 12
28.23 28.77
βˆ0 = y − βˆ1 x = − 0.924 × = 0.137
12 12
The regression model takes the form
Y = 0.924x+ 0.137 + ε
⎡ ⎛ ΣxΣy ⎞ ⎤
2

2 ⎢ ⎜ Σ xy − ⎟ ⎥
1 S xy 1 ⎢⎛ 2 (Σy ) 2 ⎞ ⎝ 12 ⎠ ⎥
(b) σˆ =
2
( S yy − )= ⎜ Σy − ⎟− = 0.123
n−2 S xx 10 ⎢⎜⎝ 12 ⎟⎠ (Σx) 2 ⎥
⎢ Σx −2
⎥
⎣⎢ 12 ⎦⎥
σˆ = 0.352
S xy
(c) r = = 0.862
[
S xx S yy
1/ 2
]
(d) 0.572, 0.961

PROBLEM SOLUTION MANUAL FOR CHAPTER 6 - PAGE 2 (OUT OF 20)

 Applied Statistics for Civil and Environmental Engineers
BY N.T. KOTTEGODA AND R. ROSSO ©

Ten steel wires of diameter 0.5 mm and length 2.5 m were extended in a
laboratory by applying vertical forces of varying magnitudes. Results are as follows
Force, kg 15 19 25 35 42 48 53 56 62 65
Increase in length, mm 1.7 2.1 2.5 3.4 3.9 4.9 5.4 5.7 6.6 7.2

(1) Estimate the parameters of a simple linear regression model with force as the explanatory variable
(2) Find 95 percent confidence limits for the two parameters
(3) Test the hypothesis that the intercept is zero
(4) What are the conclusions?

Sketch of increase in length, mm vs. force, kg

7⏐ ▪
6⏐ ▪
5⏐ ▪
4⏐ ▪
3⏐ ▪
2⏐ ▪
Force, kg 10 15 20 25 30 35 40 45 50 55 60 65 70
(a) The parameters are βˆ 0 = −0.1212 and βˆ1 = 0.1062 The regression model takes
the form
INC = 0.1062 force - 0.1212 + ε
(b) The 95% confidence limits for β1 are 0.095 and 0.118 and for
β 0 are -0.633 and 0.391
(c) Do not reject the NH that the intercept is zero
(d) The fit is good.

Table E7.2 gives 61 years of rainfall and runoff (see columns 6 and 7)
at Pontelagoscuro, on the Po river, in northeast Italy. Fit a simple regression model. Test the hypothesis
that the slope is zero. Comment on the results. Suggest methods of forming a multiple regression model
and the inclusion of other measurements that can enhance the relationship.

The parameters are βˆ 0 = 475 and βˆ1 = 0.943 The regression model takes the form
RUNOFF = 0.943 rainfall +475 + ε
The r value is 0.911
The sample t value for the slope is 17.015
Clearly, the NH of zero slope is rejected
The 95% confidence limits for β1 are 0.7077 and 0.9752
To improve the regression model, include antecedent rainfall, allow for evaporation
losses, consider the inclusion of ground ware contributions

PROBLEM SOLUTION MANUAL FOR CHAPTER 6 - PAGE 3 (OUT OF 20)

700 Applied Statistics for Civil and Environmental Engineers

Table E.7.2 Hydrologic data of the Po basin at Pontelagoscuro, northern Italy, from 1918 to 1978

Annual Annual Annual Mean

maximum flow minimum flow minimum 7-day flow annual flow Rainfall Runoff
Year (m3 /s) (m3 /s) (m3 /s) (m3 /s) (mm) (mm)
1918 5390 910 944 2010 1133 904
1919 4240 826 856 1440 999 648
1920 7220 969 1005 2400 1501 1080
1921 3000 569 590 1220 807 549
1922 2590 380 394 1070 1051 481
1923 2980 563 580 1280 969 576
1924 3920 749 813 1400 997 630
1925 3460 696 713 1530 1090 688
1926 8850 740 757 2040 1356 918
1927 3760 825 861 1630 1133 733
1928 8600 538 563 1800 1171 810
1929 2220 428 463 1090 876 490
1930 5400 607 618 1660 1159 747
1931 3700 508 527 1180 993 531
1932 4150 555 571 1420 1112 639
1933 4690 437 463 1310 1128 589
1934 6810 855 875 2050 1345 922
1935 6620 529 565 1750 1290 787
1936 6620 787 797 2310 1259 1039
1937 7700 668 675 2130 1529 958
1938 4380 287 305 1150 940 517
1939 3900 745 800 1780 1196 801
1940 5420 424 447 1350 1046 607
1941 6870 720 749 1860 1218 837
1942 4600 366 383 1160 948 522
1943 3270 310 311 987 896 444
1944 3660 306 307 905 950 407
1945 6830 304 306 916 846 412
1946 5130 655 669 1510 1011 679
1947 5460 588 594 1300 1096 585
1948 6630 711 735 1610 1100 724
1949 7220 275 278 967 922 435
1950 3260 400 411 1020 978 459
1951 8940 830 837 2200 1496 990
1952 4200 450 510 1110 913 499
1953 7400 520 550 1370 1046 616
1954 4450 440 456 1500 1100 675
1955 2400 423 441 1060 886 477
1956 5090 426 445 1220 1028 549
1957 6990 540 544 1390 1215 625
1958 5680 425 466 1320 1142 594
1959 7730 470 524 1900 1422 855
1960 6510 939 981 2620 1654 1179
1961 4880 424 450 1330 987 598
1962 4540 354 361 1070 909 481
1963 6430 732 737 1980 1362 891
1964 5630 444 461 1370 1026 616
1965 6110 321 363 1300 1015 585
ab
a b b
 Data Lists 701

Table E.7.2 (continued )

Annual Annual Annual Mean

maximum flow minimum flow minimum 7-day flow annual flow Rainfall Runoff
Year (m3 /s) (m3 /s) (m3 /s) (m3 /s) (mm) (mm)
1966 7240 359 425 1570 1228 706
1967 2470 414 454 1120 885 504
1968 7830 653 711 1650 1264 742
1969 6080 577 593 1410 995 634
1970 3170 376 394 1070 986 481
1971 5270 433 469 1380 1017 621
1972 5940 664 688 1900 1349 855
1973 4030 474 486 1270 1029 571
1974 5590 421 442 1422 959 640
1975 5360 488 508 1730 1323 778
1976 8030 323 340 1750 1318 787
1977 7800 1030 1103 2630 1564 1183
1978 5540 650 668 1920 1197 864
ab
a b b
Table E.7.3 Earthquake catalog for California north of latitude 32 and south of latitude 36◦
◦

Date and time Date and time

(day/month/year Zone (day/month/year Zone
hour minute) Magnitude typea hour minute) Magnitude typea
29/11/1852 20.00 6.5 C 08/06/1934 4.47 6.1 B
11/7/1855 4.15 6.0 B 25/02/1937 16.49 6.0 A
9/1/1857 16.00 7.8 B 19/05/1940 4.36 6.9 B
16/12/1858 10.00 6.0 C 21/10/1942 16.22 6.6 A
27/5/1862 20.00 6.0 B 15/03/1946 13.49 6.0 C
15/11/1875 22.30 6.2 B 10/04/1947 15.58 6.6 C
5/9/1883 12.30 6.2 C 04/12/1948 23.43 6.0 A
9/2/1890 12.06 6.5 A 21/07/1952 11.52 7.5 B
24/2/1892 7.20 7.0 B 21/07/1952 12.05 6.4 B
28/5/1892 11.15 6.5 A 29/07/1952 7.03 6.3 C
30/7/1894 5.12 6.0 A 22/11/1952 7.46 6.0 C
25/12/1899 12.25 6.4 A 19/03/1954 9.54 6.4 A
03/03/1901 7.45 6.4 B 28/06/1966 4.26 6.0 B
19/04/1906 0.30 6.2 B 09/04/1968 2.28 6.5 A
04/11/1908 8.37 6.0 C 09/02/1971 14.00 6.7 B
23/06/1915 3.59 6.0 B 15/10/1979 23.16 6.4 B
10/11/1916 9.11 6.1 C 02/05/1983 23.42 6.4 C
21/04/1918 22.32 6.8 A 08/07/1986 9.20 6.2 A
10/03/1922 11.21 6.1 B 24/11/1987 13.16 6.5 A
23/07/1923 7.30 6.0 A 23/04/1992 4.50 6.1 C
29/06/1925 14.42 6.9 C 28/06/1992 11.57 7.3 B
04/11/1927 13.50 7.3 B 28/06/1992 15.05 6.2 C
11/03/1933 1.54 6.2 B 17/01/1994 11.18 6.7 B
a Type A zones contain faults for which paleoseismic data suffice to estimate conditional probabilities. Type B

zones contain faults with insufficient data for conditional probability analysis. Type C zones contain diverse
or hidden faults.
Source: Data from Working Group on California Earthquake Probabilities (1995).