FACULTY OF INDUSTRIAL SCIENCES & TECHNOLOGY

FINAL EXAMINATION

COURSE : APPLIED STATISTICS

COURSE CODE : BUM2413

LECTURER : MOHD KHAIRUL BAZLI BIN MOHD AZIZ

KU MUHAMMAD NA’IM BIN KU KHALIF
NORATIKAH BINTI ABU
NORYANTI BINTI MUHAMMAD
NUR ZAHIRAH BINTI MD NOOR
ROSLINAZAIRIMAH BINTI ZAKARIA
SITI ROSLINDAR BINTI YAZIZ
SITI ZANARIAH BINTI SATARI
WAN NUR SYAHIDAH BINTI WAN YUSOFF

DATE : 03 JUNE 2018

DURATION : 3 HOURS

SESSION/SEMESTER : SESSION 2017/2018 SEMESTER II

PROGRAMME CODE : BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/

BHM/BKC/BMA/BMM/BPN/BPP/BPS/BPT/
BSB/BSK/BSP/BTC/BTE/BTM/BTP/BTV

INSTRUCTIONS TO CANDIDATES:
1. This question paper consists of SIX (6) questions. Answer ALL questions.
2. All answers to a new question should start on new page.
3. All the calculations and assumptions must be clearly stated.
4. All calculations must be in FOUR (4) decimal places.

EXAMINATION REQUIREMENTS:
Statistical Tables

DO NOT TURN THIS PAGE UNTIL YOU ARE TOLD TO DO SO

This examination paper consists of ELEVEN (11) printed pages including front page.
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

QUESTION 1

Korn Krispies is a brand of breakfast cereal and packed in boxes with a net weight of 296
grams (g). Due to high demand, the manufacturer has installed a new and more efficient
machine to fill the boxes with the cereal. However, to meet the government’s regulation,
the variance of weight of the cereal boxes should not exceed the present level of 15.
Table 1 shows the weight of a random sample of cereal boxes packed by the current and
newly installed machines. Assume that these independent samples came from a normal
population.

Table 1: Weight of Cereal for Current and New Machines

Current Machine New Machine
301.0 305.1 295.3 320.4
292.4 290.0 289.4 312.2
293.6 297.0 288.5 292.9
298.7 299.8 299.8 300.2
291.1 302.2 293.6 276.3
308.9 280.3

Figure 1 presents the Microsoft Excel output for the related statistical test on whether an
increase in the variability has occurred at 1% significance level.

F-Test Two-Sample for Variances

New Machine Current Machine

Mean 296.4833 297.0900
Variance 163.9761 26.2032
Observations 12 10
df 11 9
F 6.2579
P(F<=f) one-tail 0.0052
F Critical one-tail 5.1779

Figure 1: Microsoft Excel Output for the Related Statistical Test

2
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

(i) Conduct a necessary analysis to test the variability of the current machine that do not
exceed the present level at 2.5% significance level.
(8 Marks)

(ii) Based on Figure 1, determine whether the use of new machine has increased the
variability by using P-value approach.
(5 Marks)

(iii) In reality, the new machine did not increase the variability. Based on your answer in
(ii), identify the type of error and give a reason.
(2 Marks)

(iv) The manufacturer run an analysis for the new machine to determine whether the
standard deviation of weight of the cereal boxes is equal to the present level and
found that the 90% confidence interval is ( 6.6386,13.2772) g . Using the given

confidence interval, can we conclude that the standard deviation in the packed weight
is equal to the present level?
(4 Marks)

QUESTION 2

A laboratory experiment is conducted to investigate the effect of an additive and

temperatures to the adhesiveness of rubber product. Nine products were made with the
additive and another nine without additive. The adhesive strength (in psi) of the products
is measured based on three level of temperatures: low (20 0C ) , moderate (60 0C ) and high

(90 0C ) . Based on the assumption that the adhesiveness is normally distributed, the analysis
of variance test is conducted on the data. The incomplete corresponding Microsoft Excel
output is presented in Table 2.

3
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

Table 2: Incomplete Microsoft Excel Output of ANOVA Table

Source of
Variation SS df MS F P-value F crit
Additive 12.6672 3.6E-09 4.7472
Temperature 7.0933 2 3.5467 63.84 4.02E-07 3.8853
Interaction 0.5378 2 4.84 0.0288 3.8853
Within 0.6667 12 0.0556

Total 20.9650 17

(i) Identify the dependent variable.

(1 Mark)

(ii) Name the factor(s) involved.

(2 Marks)

(iii) How many treatment(s) involved?

(1 Mark)

(iv) How many replications for each treatment?

(1 Mark)

(v) Find the f-test for the sample factor in Table 2.

(2 Marks)
(vi) Investigate the effect of the additive and temperatures to the adhesiveness of rubber
product by using P-value approach.

(6 Marks)

4
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

QUESTION 3

Alia sells life insurance for the Prudence Insurance Company. She sells insurance by
making visits to her clients’ homes. Alia believes that the number of sales depends on the
number of home visits made. For the past several years, she has kept careful records of the
number of visits and the number of people who buys insurance each week. For a random
sample of ten weeks, the data is given in Table 3.

Table 3: Weekly Records of Insurance Sale

Number of visit Number of people who buys insurance
11 3
19 11
16 8
13 5
28 8
5 2
20 5
14 6
22 8
7 3

(i) Identify the predictor and response variables.

(1 Mark)

(ii) Calculate the correlation coefficient and interpret its value.

(8 Marks)

(iii) Based on the value of correlation coefficient in (ii), does it make sense to use the
linear regression model for prediction purposes? Give a reason.
(2 Marks)

5
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

(iv) What is the percentage of the variation in response variable that can be explained by
the corresponding predictor variable?
(2 Marks)

(v) Based on the following Microsoft Excel output, test the linearity between the
predictor and response variables at 97% confidence interval by using P-value
approach.

ANOVA
df SS MS F Significance F
Regression 1 41.4921 41.4921 10.5686 0.0117
Residual 8 31.4079 3.9260
Total 9 72.9

(5 Marks)

(vi) If Alia makes 18 visits in a week, predict how many people will buy insurance from
her?
(8 Marks)

(vii) If Alia uses the estimated linear regression model in (vi) to predict the number of
people who buys insurance in that week for 50 number of visits, did Alia do the
extrapolation or interpolation for the value? Give one implication of such prediction.
(2 Marks)

(viii) Based on the estimated linear regression model in (vi), what is the marginal change
in the estimated number of people who buys insurance for each number of visits?
(1 Mark)

6
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

QUESTION 4

The number of weekly motorcyclist is affected by three variables;

x1 : Price of fuel per week
x 2 : Monthly income of riders

x 3 : Average parking rates per month

An analysis of relationship between the number of weekly motorcyclist with the predictor
variables yields the Microsoft Excel output in Figure 2.

SUMMARY OUTPUT

Regression Statistics
Multiple R 0.9716
R Square 0.9441
Adjusted R Square 0.9368
Standard Error 5356.4047
Observations 27

ANOVA
df SS MS F Significance F
Regression 3 11137260774.9410 3712420258.3137 129.3929 0.0000
Residual 23 659894628.9109 28691070.8222
Total 26 11797155403.8519

Coefficients Standard Error t Stat P-value Lower 95% Upper 95%

Intercept 203380.1517 5009.8819 40.5958 0.0000 193016.4214 213743.8821
Price per week -719.7651 85.8945 -8.3796 0.0000 -897.4514 -542.0789
Monthly income of riders -2.1608 1.1589 -1.8645 0.0751 -4.5581 0.2366
Average parking rates per month 154.0198 73.1519 2.1055 0.0464 2.6936 305.3461

Figure 2: Microsoft Excel Outputs for Three Predictors

(i) Is there any evidence to conclude that at least one of the independent variables is
related to the dependent variable?
(5 Marks)

(ii) Based on the P-value in the coefficients table, identify the significant predictors.
(4 Marks)

7
 CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

(iii) The summary of the multiple linear regression analysis for the study is given by
Table 4. Based on Table 4, determine the best regression model to predict the
number of weekly motorcyclist and give your reason.

Table 4: Summary Table for Multiple Linear Regression

Adjusted Regression Equation
Predictor(s) P-value r2
r2
x1 3.4134 ´10-16 0.9331 0.9304 yˆ = 197208.3412 - 744.07487x1
x2 2.9544 ´10-9 0.7615 0.7520 yˆ = 220217.5532 - 5.4407x2
x3 7.9369 ´10-7 0.6295 0.6147 yˆ = 199919.4741 - 372.7065x3
x1 and x2 7.7790 ´10-15 0.9333 0.9278 yˆ = 198354.6756 - 721.7501x1 - 0.2074x2
x1 and x3 5.0805 ´10-15 0.9356 0.9302 yˆ = 195528.4138 - 807.2254 x2 - 44.8362 x3
x2 and x3 1.8436 ´10-8 0.7733 0.7544 yˆ = 225399.8027 - 7.4644 x2 + 160.7479 x3
x1 , x2 and x3 0.0000 0.9441 0.9368 yˆ = 203380.1517 - 719.7651x1 - 2.1608x2 + 154.0198x3

(3 Marks)

(iv) In your opinion, why the model selection process is needed? State a reason.
(1 Mark)

8
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

QUESTION 5

A toy company has introduced a new toy car. For quality control purposes, the defective
toy cars are rejected from the production line. Previous record shows that the probability
of defective toy cars from the production line is 0.15. In every week of year 2017, five toy
cars are randomly selected. Table 5 shows the number of weeks in which 0, 1, … , 5
defective toy cars are rejected in 2017.

Table 5: Number of Defective Toy Cars

No. of defective toy cars 0 1 2 3 4 5

No. of weeks 19 21 8 3 1 0

The quality inspector claims that the number of defective toy cars in 2017 follows a
binomial distribution.

(i) Identify the parameter value(s) used by the quality inspector.

(1 Mark)

(ii) Investigate the quality inspector’s claim at a = 0.01.

(12 Marks)

QUESTION 6

9
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

(a) Give one purpose of the goodness of fit test.

(1 Mark)

(b) In UMP, students’ intakes are enrolled from different entry qualification
backgrounds such as STPM, Diploma, Matriculation and STAM. A study is
conducted to investigate the relationship between students’ gender and entry
qualification. Table 6 gives the data collected from intake 2011 to 2016. Table 7
presents the expected frequency for students’ gender and entry qualification.

Table 6: Entry Qualification and Gender

Entry Qualification

Gender STPM Diploma Matriculation STAM

Female 785 586 2702 16

Male 972 736 1716 18

Table 7: Expected Frequency for Gender and Entry Qualification

Entry Qualification

Gender STPM Diploma Matriculation STAM

Female 953.9733 X 2398.7786 18.4605

Male 803.0267 604.2125 2019.2214 15.5395

10
CONFIDENTIAL BAA/BCG/BCN/BCS/BEE/BEP/BFF/BFM/BHA/BHM/
BKC/BMA/BMM/BPN/BPP/BPS/BPT/BSB/BSK/BSP/
BTC/ BTE/BTM/BTP/BTV/1718II/BUM2413

(i) Calculate the value of X in Table 7.

(2 Marks)

(ii) Based on Table 6 and Table 7, give a comment which relates to students with
matriculation entry qualification.
(1 Mark)

(iii) At 10% level of significance, determine whether the entry qualification is

independent of gender?

(9 Marks)

END OF QUESTION PAPER

11