Examination Answer (Semester B, 2015‐16)
Question 1 (25 marks)
a. 19‐21
Normal distribution is bell‐shaped and symmetric, with mean equals to the mode,
whereas the mode is the value which most frequently occurred. So, the range including
the mode is most likely to happen.
b. P(X > 30) = P (Z > )
= P (Z > 2.0)
= 1 ‐ P (Z 2.0)
= 1 ‐ 0.9772
= 0.0228
c. Let X = number of free pizza
P(X 1) = 1 ‐ P(X = 0)
!
=1‐ 0.0228 1 0.0228
! !
= 1 ‐ 0.8911
= 0.1089
d. P( <X< ) = 0.60
P( <Z< ) = 0.60
For symmetric distribution of probability (0.6) on both sides of ,
P(Z < ‐0.84) = 0.2 and P(Z < 0.84) = 0.8
Therefore, = ‐0.84 = ‐0.84 5 + 20 = 15.8
= 0.84 = 0.84 5 + 20 = 24.2
e. As X ~ N(20,52), ~ N(20, )
P(18 23)
= P( )
√ √
= P(‐2.0396 3.0594)
= 0.99889 – 0.0207
= 0.9782
f. P( > + 2) < 0.05
1 ‐ P( + 2) < 0.05
P( + 2) > 0.95
P(Z ) > 0.95
/√
√
1.645
n 67.6506 68 (round up)
Question 2 (25 marks)
a. min = 20 Q1 = 24 Q2 = 31 Q3 = 34 max = 47
The sample data set is right‐skewed.
b. Normal population assumption is needed. This is because the sample is drawn from an
unknown population distribution, and n = 10 <30, hence Central Limit Theorem is not
applicable
With known σ = 7, Z‐distribution is used in conducting inferential analysis
c. 95% CI for = 30.8 1.96 = [26.4614, 35.1386] μg/l
√
It is 95% confidence that the mean DMS odor threshold among all oenologists is
between 26.4614 and 35.1386 ug/l.
d. H0: μ ≤ 25 H1: μ > 25
At = 0.05, reject H0 if Z > 1.645
.
Z= = 2.6202
√
Reject H0. There is sufficient evidence that the mean odor threshold for oenologists is
higher than the published threshold, 25μg/l
e. p‐value = P(Z 2.6202) = 1 ‐ 0.9956 = 0.0044
Question 3 (25 marks)
a. P(F and O)
= P(O) ‐ P(M and O)
= P(O) ‐ P(O|M) P(M)
= 0.34 ‐ 0.35 0.76
= 0.074
b. P(M | O)
= P(M and O) / P(O)
= P(O|M) P(M) / P(O)
= 0.35 0.76 / 0.34
= 0.7824
c. P(F and O) = 0.074
P(F) P(O) = 0.24 0.34 = 0.0816
As P(F) P(O) P(F and O), overweighted and gender is not statistically independent (or they are
dependent).
‐‐ any other reasonable method showing the two events are not independent also acceptable
d. n = 500 > 30
np = 500*0.34 = 170 > 5
n(1‐p) = 500*(1 ‐ 0.34) = 330 > 5
Sampling distribution of p is approximately normal, ~ 0.3,
. ∗ .
. .
85% CI for π = = 0.34 1.44 = [0.3095, 0.3705]
We are 85% confidence that the population proportion of overweighted members are between
0.3095 and 0.3705.
e. When conducting a hypothesis test, should be used to compute the standard error. When
0.3, which is smaller than p = 0.34, the standard error decreases, the confidence interval
becomes narrower, or the Z‐test statistics becomes more extreme, leading to a higher chance to
reject the null hypothesis. Concluding there is sufficient evidence that the population proportion
is not 0.3.
. . .
f.
.
1518.08 1519
Question 4(25 marks)
a. 0.5802 15.0352
When the number of copiers serviced increased by one, the service time is predicted to increase
by 15.0352 minutes.
b. √ √0.9575 0.9785
There is a very strong positive linear relaionship between variable and .
c. Covariance depends on the the units used to measure and and thus usually cannot be
directly compared for different variables. Correlation coefficient is a “standardized score” of the
covariance and it is unit‐free
d. 95.75% of the vairation in Y has been explained by the estiamated regression equation.
e. H0: 0 H 1: > 0
At = 0.05, reject H0 if t > 1.6811 with df = 43
.
31.1223
.
Since t = 31.1223 > 1.6811, we reject H0.
There is sufficient evidence that there is a positive relationship between Y and X.
f. (i) 0.5802 15.0352 1 14.455 minutes
(ii) 0.5802 15.0352 5 74.5958 minutes
g. Part (ii)’s prediction is more justifiable because the value of X = 5 lies within the observed range.
