ANALISA STATISTIKA DAN PROBABILITAS – RM185101

Assignment 2 Hypothesis Thesis

Created By:

Ausa Ramadhan Agustawijaya (03311942000008)

Dosen Pengampu: - Ira Mutiara Anjasmara, S.T., M.Phil., Ph.D.

DEPARTEMEN TEKNIK GEOMATIKA

FAKULTAS TEKNIK SIPIL, PERENCANAAN, DAN KEBUMIAN INSTITUT

TEKNOLOGI SEPULUH NOPEMBER SURABAYA

2022

1
 ASSIGNMENT

APPLIED STATISTICS AND PROBABILITY

1. Use the 𝜒2-distribution table to determine the values of 𝜒2𝛼 that would be used to construct confidence intervals for
⁄2
a population variance for the following combinations:
(a) 𝛼 = 0.10 , 𝜈 = 25
(b) 𝛼 = 0.05 , 𝜈 = 15
(c) 𝛼 = 0.05 , 𝜈 = 10
(d) 𝛼 = 0.01 , 𝜈 = 30
Answer :
Determining CI value from X2 using :
𝑣𝑠 2 2
𝑣𝑠 2
2 ≤ 𝜎 ≤ 2
𝑥𝑣,𝑎/2 𝑥𝑥,1−𝑎/2
So, by the formula we can determine :
a) 𝛼 = 0.10, v = 25
Confidence level = 1- 𝛼 = 90%
2 2
𝑥𝑣,𝑎/2 = 𝑥25,0.05 = 37.653
2 2
𝑥𝑣,𝑎/2 = 𝑥25,0.95 = 14.611
25𝑠 2 25𝑠 2
≤ 𝜎2 ≤
37.653 14.611
b) 𝛼 = 0.05, v = 15
Confidence level = 1- 𝛼 = 95%
2 2
𝑥𝑣,𝑎/2 = 𝑥15,0.025 = 27.488
2 2
𝑥𝑣,𝑎/2 = 𝑥15,0.975 = 6.262
15𝑠 2 2
15𝑠 2
≤𝜎 ≤
27.488 6.262
c) 𝛼 = 0.05, v = 10
Confidence level = 1- 𝛼 = 95%
2 2
𝑥𝑣,𝑎/2 = 𝑥10,0.025 = 20.483
2 2
𝑥𝑣,𝑎/2 = 𝑥10,0.975 = 3.247
10𝑠 2 10𝑠 2
≤ 𝜎2 ≤
20.483 3.247
d) 𝛼 = 0.05, v = 30
Confidence level = 1- 𝛼 = 99%
2 2
𝑥𝑣,𝑎/2 = 𝑥30,0.005 = 53.672
2 2
𝑥𝑣,𝑎/2 = 𝑥30,0.995 = 13.787
30𝑠 2 30𝑠 2
≤ 𝜎2 ≤
53.672 13.787

2. Use the 𝑡-distribution table (Table D.2) to determine the values of 𝑡𝛼⁄2 that would be used to construct confidence
intervals for a population variance for the following combinations:
(a) 𝛼 = 0.10 , 𝜈 = 25
(b) 𝛼 = 0.05 , 𝜈 = 15
(c) 𝛼 = 0.01 , 𝜈 = 10
(d) 𝛼 = 0.01 , 𝜈 = 40
Answer :
a) 𝛼 = 0.10 , 𝜈 = 25
Confidence level = 1- 𝛼 = 90%
𝑡𝑣,𝑎/2 = 𝑡25,0.05 = 1.708
2
 𝑠
𝐶𝐼 = x̅ ± 1.708 ×
√𝑛
𝑠 𝑠
x̅ − (1.708 × ) ≤ 𝜇 ≤ x̅ + (1.708 × )
√𝑛 √𝑛
b) 𝛼 = 0.05 , 𝜈 = 15
Confidence level = 1- 𝛼 = 95%
𝑡𝑣,𝑎/2 = 𝑡15,0.025 = 2.131
𝑠
𝐶𝐼 = x̅ ± 2.131 ×
√𝑛
𝑠 𝑠
x̅ − (2.131 × ) ≤ 𝜇 ≤ x̅ + (2.131 × )
√𝑛 √𝑛
c) 𝛼 = 0.01 , 𝜈 = 10
Confidence level = 1- 𝛼 = 99%
𝑡𝑣,𝑎/2 = 𝑡10,0.005 = 3.169
𝑠
𝐶𝐼 = x̅ ± 3.169 ×
√𝑛
𝑠 𝑠
x̅ − (3.169 × ) ≤ 𝜇 ≤ x̅ + (3.169 × )
√𝑛 √𝑛
d) 𝛼 = 0.01 , 𝜈 = 40
Confidence level = 1- 𝛼 = 99%
𝑡𝑣,𝑎/2 = 𝑡40,0.005 = 2.704
𝑠
𝐶𝐼 = x̅ ± 2.704 ×
√𝑛
𝑠 𝑠
x̅ − (2.704 × ) ≤ 𝜇 ≤ x̅ + (2.704 × )
√𝑛 √𝑛
3. Use the 𝐹-distribution table (Table D.2) to determine the values of 𝐹𝛼,𝑣1,𝑣2 that would be used to construct confidence
intervals for a population variance for the following combinations:
(a) 𝛼 = 0.20 , 𝜈1 = 24, 𝜈2 = 2
(b) 𝛼 = 0.01 , 𝜈1 = 24, 𝜈2 = 2
(c) 𝛼 = 0.05 , 𝜈1 = 60, 𝜈2 = 20
(d) 𝛼 = 0.80 , 𝜈1 = 2, 𝜈2 = 24

Answer :
a) 𝛼 = 0.20 , 𝜈1 = 24, 𝜈2 = 2
𝐹𝑎,𝑣1,𝑣2 = 𝐹0.20,24,2 = 4.439882584
b) 𝛼 = 0.01 , 𝜈1 = 24, 𝜈2 = 2
𝐹𝑎,𝑣1,𝑣2 = 𝐹0.01,24,2 = 99.45750162
c) 𝛼 = 0.05 , 𝜈1 = 60, 𝜈2 = 20
𝐹𝑎,𝑣1,𝑣2 = 𝐹0.05,60,20 = 1.95
d) 𝛼 = 0.80 , 𝜈1 = 2, 𝜈2 = 24
𝐹𝑎,𝑣1,𝑣2 = 𝐹0.80,2,20 = 0.22531181

4. A least squares adjustment is computed twice on a data set. When the data are minimally constrained with 10
degrees of freedom, a variance of 1.07 is obtained. In the second run, the fully constrained network has 12 degrees
of freedom with a standard deviation of 1.53. The a priori estimates for the reference variances in both adjustments
are 1; that is, 𝜎12 = 𝜎22 = 1.
(a) What is the 95% confidence interval for the ratio of the two variances? Is there reason to be concerned
about the consistency of the control? Justify your response statistically.
(b) What is the 95% confidence interval for the reference variance in the minimally constrained
adjustment? The population variance is 1. Does this interval contain 1?
(c) What is the 95% confidence interval for the reference variance in the fully constrained adjustment? The
population variance is 1. Does this interval contain 1?
Answer :
a) 95% confidence
𝑉𝐴 = 10
𝑆𝐴2 = 1.07

3
𝑉𝐵= 12
𝑆𝐵= 1.53
𝑆𝐵2
= 2.3409
- Formulate Hypothesis
Ha = 𝜎𝐴2 ≠ 𝜎𝐵2
Ho = 𝜎𝐴2 = 𝜎𝐵2
- Two tailed test
- Significant level and degree of freedom
α = 0.05
𝑉𝐴 = 10 (Denominator)
𝑉𝐵 = 12 (Denominator)
- 𝐹𝑉𝐵,𝑉𝐴,𝛼/2 = 𝐹12,10,0.025 = 3.62
1 1
- 𝐹𝑉𝐵,𝑉𝐴,−𝛼/2 = 𝐹12,10,0.975 = 𝐹 = 3.37
= 0.2967
12,10,0.975
- Rejection region

- F-score
𝑆𝐵2 2.3409
𝐹= 2= = 2.1878
𝑆𝐴 1.07
- Comparing the test statistic
0.2967 < 2.1878 < 3.62 , therefore F does not lie in the rejection region. Hence, we
do not reject H0 at significant level.
- No significant difference between two variances aat 95% confidence.
b) For the reference variance in the minimally constrained adjustment
𝑉𝐴 = 12
𝑆𝐴2 = 1.07
- Formulate Hypothesis
Ha = 𝜎𝐴2 ≠ 1
Ho = 𝜎𝐴2 = 1
- Two tailed test
- Level of significant and degree of freedom
α = 0.05
v = 10
- Critical Value
2 2
𝑋𝑣,𝑎/2 = 𝑋10,0.025 = 20.4831
2 2
𝑋𝑣,1−𝑎/2 = 𝑋10,0.975 = 3.24697
- Confidence interval
𝑉𝑆 2 2
𝑉𝑆 2
𝐶𝐼 = 2 ≤ 𝜎 ≤ 2
𝑋𝑣,𝑎/2 𝑋𝑣,1−𝑎/2
10 × 1.07 10 × 1.07
𝐶𝐼 = ≤ 1≤
20.4831 3.24697
𝐶𝐼 = 0.52238 ≤ 1 ≤ 3.29538

- Showing interval region contain 1 as the populaton variance

- The sample variance does lie within this interval. Hence, we don’t reject H0 with
95% confidence. Our sample is compatible with supposed population variance at
0.05 ignificant.
4
 c) Reference variance fully constrained
𝑉𝐴 = 12
𝑆𝐵2 = 2.3409
- Formulate Hypothesis
Ha = 𝜎𝐴2 ≠ 1
Ho = 𝜎𝐴2 = 1
- Two tailed test
- Level of significant and degree of freedom
α = 0.05
v = 12
- Critical value
2 2
𝑋𝑣,𝑎/2 = 𝑋12,0.025 = 23.3367
2 2
𝑋𝑣,1−𝑎/2 = 𝑋12,0.975 = 4.40379
- Confident Interval
𝑉𝑆 2 𝑉𝑆 2
𝐶𝐼 = 2 ≤ 𝜎2 ≤ 2
𝑋𝑣,𝑎/2 𝑋𝑣,1−𝑎/2
12 × 2.3409 12 × 2.3409
𝐶𝐼 = ≤ 𝜎2 ≤
23.3367 4.40379
𝐶𝐼 = 1.2037 ≤ 𝜎 2 ≤ 6.3788

- Determine the rejection region

- Region shows 𝜎 2 = 1 is out of confidence interval

- Hence, we reject H0 with 95% confidence. The sample is incompatible with
supposed population variance at 0.05 significant

5. The calibrated length of a baseline is 402.167 m. An average distance of 402.151 m with a standard deviation of
±0.0055 m is computed after the line is observed five times with an EDM.
(a) What is the 95% confidence interval for the measurement?
(b) At a 95% level of confidence, can you state that the EDM is working properly? Justify your response
statistically.
(c) At a 90% level of confidence can you state that the EDM is working properly? Justify your response
statistically.
Answer :
a) 95% confidence nterval for the measurement
𝜇 = 402.167m
x̅ = 402.151 m
𝑠 = 0.0055m
𝑛=5
𝑣 =5−1=4
α = 0.05
5
 - Determine t-value

t0.025=2.776
- Standard error
Sx = 0.00246
- CI = 𝑥 ± 𝑡𝑣,𝛼/2. 𝑠𝑥
CI = 402.14417 𝑡𝑜 402.15783
b) 95% level confidence, wether EDM is working properly or not
Since 402.167 does not lie within the confidence interval, we can not say with
confidence that the EDM is working properly
c) At a 90% level of confidence can you state that the EDM is working properly?
Justify your response statistically.
α = 0.1 --> α/2 = 0.05
t0.05=2.132
ν = n-1 = 5-1 = 4
0.0055 0.0055
- 402.151 - 2.132 < µ < 402.151 + 2.132
√5 √5

402.146 < µ < 402.156

Since 402.167, does not lie within interval we still can not say with confidence that the EDM
is working properly

6. A mean length of 1023.573 m with a standard deviation of ±0.0056 m is obtained for a distance after five
observations. Using the technical specifications, it is found that the standard deviation for this observation should be
±0.0043 m.
(a) Perform a statistical test to check the repeatability of the instrument at a level of significance of 0.05.
(b) Perform a statistical test to check the repeatability of the instrument at a level of significance of 0.01.
Answer :

(a) 𝑥 = 1023.573 𝑚
𝑠 = ± 0.0056 𝑚
𝜎 = ± 0.0043
𝜎 2 = 1.849 × 10−5
Check the repeatability of the instrument at a level of significance of 0.05 :
• Formulate hypothesis
𝐻1 = 𝜎 2 ≠ 1.849 × 10 −5 (Alternative hypothesis)
𝐻0 = 𝜎 2 = 1.849 × 10 −5 (Alternative hypothesis)
• There are 2 tailed test
• Level of significant and degree of freedom
𝛼 = 0.05 → 1 – 𝛼 = 95%
𝑛 = 5
𝑣 = 𝑛−1 = 4
• Critical value (𝑋 2 )
𝑋 2 𝑣,𝛼/2 = 𝑋 2 4,0.025 = 11.1433
𝑋 2 𝑣,1−𝛼/2 = 𝑋 2 4,0.975 = 0.484419

- Statistical test (𝑋 2 − 𝑠𝑐𝑜𝑟𝑒)

6
 𝑉.𝑆 2 4.0.00562
𝑋 2 − 𝑠𝑐𝑜𝑟𝑒 = = = 6.78420768 = 6.784
𝜎2 0.00432

- Compare the statistical test against the critical value

0.484 < 6.784 < 11.143, don’t lie in rejection region.
Therefore, we do not reject H0 at the 0.05 significant level.
- Our sample is compatible with supposed population mean at 0.05 level of
significant. Therefore it follows that the observations agree with result using
technical spesifications.
•
(b) 𝑥 = 1023.573 𝑚
𝑠 = ± 0.0056 𝑚
𝜎 = ± 0.0043
𝜎 2 = 1.849 × 10−5
Check the repeatability of the instrument at a level of significance of 0.01 :
• Formulate hypothesis
𝐻1 = 𝜎 2 ≠ 1.849 × 10 −5 (Alternative hypothesis)
𝐻0 = 𝜎 2 = 1.849 × 10 −5 (Alternative hypothesis)
• There are 2 tailed test
• Level of significant and degree of freedom
𝛼 = 0.05 → 1 – 𝛼 = 95%
𝑛 = 5
𝑣 = 𝑛−1 = 4
• Critical value (𝑋 2 )
𝑋 2 𝑣,𝛼/2 = 𝑋 2 4,0.005 = 14.8602
𝑋 2 𝑣,1−𝛼/2 = 𝑋 2 4,0.995 = 0.206990

• Rejection region

• Statistical test (𝑋 2 − 𝑠𝑐𝑜𝑟𝑒)

7
 𝑉.𝑆 2 4.0.00562
𝑋 2 − 𝑠𝑐𝑜𝑟𝑒 = = = 6.78420768 = 6.784
𝜎2 0.00432

• Compare the statistical test against the critical value

0.206990 < 6.784 < 14.8602, do not lie in rejection region.
Therefore, we do not reject H0 at the 0.01 significant level.
• Our sample is compatible with supposed population mean at 0.01 level of
significant. Therefore it follows that the observations agree with result using
technical spesifications.

7. A least squares adjustment is computed twice on a data set. When the data are minimally constrained with 24
degrees of freedom, a reference variance of 0.89 is obtained. In the second run, the fully constrained network, which
also has 24 degrees of freedom, has a reference variance of 1.15. The a priori estimate for the reference variance in
both adjustments is 1; that is, 𝜎12 = 𝜎22 = 1.
(a) Are the two variances statistically equal at a 0.05 level of significance?
(b) Is the minimally constrained adjustment reference variance statistically equal to 1 at a 0.05 level of
significance?
(c) Is the fully constrained adjustment reference variance statistically equal to 1 at a 0.05 level of
significance?
(d) Is there a statistical reason to be concerned about the presence of errors in either the control or the
observations?

Answer :

(a) Compare two variances :

𝑉𝐴 = 24

𝑆𝐴2 = 0.89

𝑉𝐵 = 24

𝑆𝐵2 = 1.15
• Formulate hyphothesis
𝐻1 = 𝜎𝐴 2 ≠ 𝜎𝐵 2
𝐻0 = 𝜎𝐴 2 = 𝜎𝐵 2
• There are two tailed test
• Significant level and degree of freedom
𝛼 = 0.05
𝑉𝐴 = 24 (Denominator)
𝑉𝐵 = 24 (Numerator)
Because of 2 tailed test, so :
𝐹𝑉𝐵,𝑉𝐴,𝛼/2 =𝐹24,24,0.025 = 2.27
1 1
𝐹VB,VA,1−α/2 =𝐹24,24,0.975 =𝐹 = 2.27 = 0.4405
24,24,0.025

• Rejection region

8
 The rejection region is F < 0.4405 and F > 2.27
• Determine F-score
S 2 1.15
F-score = S𝐴 2 = 0.89 = 1.292134 = 1.292
𝐵

• Comparing the statistical test

0.4405 < 1.2921 < 2.27, therefore F does not lie in the rejection region.
Therefore, we do not reject H0 at the 0.05 significant level
• There is no significant difference between two variances at 95% confidence.
Therefore, two variances statistically equal at 0.05 level of significant.
(b) The minimally constrained
𝑉𝐴 = 24

𝑆𝐴2 = 0.89

Compare two variances :

• Formulate hyphothesis
𝐻1 = 𝜎𝐴 2 ≠ 1
𝐻0 = 𝜎𝐴 2 = 1
• There are two tailed test
• Significant level and degree of freedom
𝛼 = 0.05
V = 24
• Critical value
𝑋 2 𝑣,𝛼 = 𝑋 2 24,0.025 = 39.3641
2

𝑋 2 𝑣,1−𝛼/2 = 𝑋 2 24,0.975 = 12.4011

• Confident Interval
𝑉. 𝑆 2 2
𝑉. 𝑆 2
𝐶𝐼 = ≤ 𝜎 ≤
𝑋 2 𝑣,𝛼/2 𝑋 2 𝑣,1−𝛼/2

9
 24 × 0.89 24 × 0.89
𝐶𝐼 = ≤1≤
39.3641 12.4011
𝐶𝐼 = 0.5426 ≤ 1 ≤ 1.7224
• Rejection region
Because the null hypothesis will be rejected if 𝜎𝐴 2 ≠ 𝜎𝐵 2 , since the test of 𝜎𝐴 2 ≠
𝜎𝐵 2 by using 𝑋 2 . So, the rejected areas are :
𝑋 2 < 0.5426 𝑎𝑛𝑑 𝑋 2 > 1.7224
• Comparing the statistical test
0.5426 ≤ 1 ≤ 1.7224, therefore F does not lie in the rejection region.
Therefore, we do not reject H0 at the 0.05 significant level
• There is no significant difference between two variances at 95% confidence.
Therefore, two variances statistically equal at 0.05 level of significant.
(c) Fully constrained
𝑉𝐵 = 24
𝑆𝐵2 = 1.15
• Formulate hyphothesis
𝐻1 = 𝜎𝐵 2 ≠ 1
𝐻0 = 𝜎𝐵 2 = 1
• There are two tailed test
• Significant level and degree of freedom
𝛼 = 0.05
V = 24
• Critical value
𝑋 2 𝑣,𝛼 = 𝑋 2 24,0.025 = 39.3641
2
2
𝑋 𝑣,1−𝛼/2 = 𝑋 2 24,0.975 = 12.4011
• Confident Interval
𝑉. 𝑆 2 2
𝑉. 𝑆 2
𝐶𝐼 = 2 ≤𝜎 ≤ 2
𝑋 𝑣,𝛼/2 𝑋 𝑣,1−𝛼/2
24 × 1.15 24 × 1.15
𝐶𝐼 = ≤1≤
39.3641 12.4011
𝐶𝐼 = 0.7011 ≤ 1 ≤ 2.2256
• Rejection region
Because the null hypothesis will be rejected if 𝜎𝐴 2 ≠ 𝜎𝐵 2 , since the test of 𝜎𝐴 2 ≠
𝜎𝐵 2 by using 𝑋 2 . So, the rejected areas are :
𝑋 2 < 0.7011 𝑎𝑛𝑑 𝑋 2 > 2.2256
• Comparing the statistical test
10
 0.7011 ≤ 1 ≤ 2.2256, therefore F does not lie in the rejection region.
Therefore, we do not reject H0 at the 0.05 significant level
• There is no significant difference between two variances at 95% confidence.
Therefore, two variances statistically equal at 0.05 level of significant.
(d) In the answer of point (b) and (c), we compare the minimally and fully constrained
adjustment reference variance statistically equal to 1 and the statistics result is equal
to 1. Maybe it can work if the population variance is 1, so the interval can still
contain 1. There is no reason to be concerned about the presence of errors in either
the control or the observations.

8. An EDM is placed on a calibration baseline and the distance between two monuments is determined to be 1200.012
m ± 0.047 m after 10 observations. The length between the monuments is calibrated as 1200.005 m. Is the
instrument measuring the length properly at:
(a) a 0.01 level of significance?
(b) a 0.05 level of significance?
(c) a 0.10 level of significance?

Answer :
𝜇 = 1200.005 𝑚
𝑥 = 1200.012 𝑚
𝑠 = 0.047 𝑚
𝑛 = 10
(a) 0.01 level of significance
𝛼 = 0.01
• Formulate hyphothesis
𝐻1 = 𝜇 ≠ 1200.005
𝐻0 = 𝜇 = 1200.005
• There are two tailed test
• Level of significant and degree of freedom
𝛼 = 0.01
𝑛 = 10 → 𝑣 = 𝑛 − 1 = 9
• Critical value
𝑡v,α/2 = 𝑡9,0.005 = 3.250
• Rejection region

11
 The rejection region is t > 3.250 and t < -3.250
• Determine t-score
𝑥 − 𝜇 1200.012 − 1200.005
𝑡 − 𝑠𝑐𝑜𝑟𝑒 = = = 0.47098 = 0.471
𝜎𝑥 0.047⁄
√10
• Compare the statistical test agains the critical value
-3.250 < 0.471 < 3.250. So, we do not reject H0 at the 0.01 significant level.
• This measurement compatible with calibrated length of a distance. It’s means
that the instrument measuring the length properly at 0.01 level of significance
(b) 0.05 level of significance
𝛼 = 0.05
• Formulate hyphothesis
𝐻1 = 𝜇 ≠ 1200.005
𝐻0 = 𝜇 = 1200.005
• There are two tailed test
• Level of significant and degree of freedom
𝛼 = 0.05
𝑛 = 10 → 𝑣 = 𝑛 − 1 = 9
• Critical value
𝑡v,α/2 = 𝑡9,0.025 = 2.262
• Rejection region

12
 The rejection region is t > 2.262 and t < -2.262
• Determine t-score
𝑥 − 𝜇 1200.012 − 1200.005
𝑡 − 𝑠𝑐𝑜𝑟𝑒 = = = 0.47098 = 0.471
𝜎𝑥 0.047⁄
√10
• Compare the statistical test agains the critical value
-2.62 < 0.471 < 2.262. So, we do not reject H0 at the 0.05 significant level.
• This measurement compatible with calibrated length of a distance. It’s means
that the instrument measuring the length properly at 0.05 level of significance
(c) 0.10 level of significance
𝛼 = 0.10
• Formulate hyphothesis
𝐻1 = 𝜇 ≠ 1200.005
𝐻0 = 𝜇 = 1200.005
• There are two tailed test
• Level of significant and degree of freedom
𝛼 = 0.10
𝑛 = 10 → 𝑣 = 𝑛 − 1 = 9
• Critical value
𝑡v,α/2 = 𝑡9,0.05 = 1.833
• Rejection region

13
The rejection region is t > 1.833 and t < -1.833
• Determine t-score
𝑥 − 𝜇 1200.012 − 1200.005
𝑡 − 𝑠𝑐𝑜𝑟𝑒 = = = 0.47098 = 0.471
𝜎𝑥 0.047⁄
√10
• Compare the statistical test agains the critical value
-1.833< 0.471 < 1.833. So, we do not reject H0 at the 0.10 significant level.
• This measurement compatible with calibrated length of a distance. It’s means
that the instrument measuring the length properly at 0.10 level of significance

14