STAT 404 - Assignment 3 Solutions

Total marks: 60
Due date: Saturday, Nov. 4 at 11:59pm

Reminder:
• Follow the Assignment Guidelines.
• Select pages when submitting to Gradescope.
• One-line R commands should only be used to verify answers.
• When performing a test, state the hypotheses, test statistic
and distribution, p-value or critical value, and conclusion. If
unspecified, use the default 5% significance level.

1. [20] We are interested in whether students do better on the midterm

or on the final exam. The marks of 20 students from some course are
given below. We wish to answer this question by performing a test
for H0 : δ = 0 against the alternative H1 : δ ̸= 0 based on the paired
observations. Assume common notation and model assumptions.

midterm = c(82.54,79.37,77.78,60.32,65.08,69.84,61.90,61.90,44.44,
63.49,80.95,84.13,76.19,88.89,73.02,85.71,71.43,84.13,84.13,69.84)
final = c(85.67,84.44,80.83,69.06,67.73,81.75,73.62,62.61,66.56,
65.52,82.38,94.13,65.58,96.68,86.40,87.95,74.92,84.56,85.99,57.34)

Remark: this group of students may not be representative of the general

student population, nor may this course be representative of all courses
at UBC or other places. The data set is used for illustration only.

(a) [4] Perform the paired t-test for the above data.
Answer. The hypotheses are stated in the question. We first find
the differences for each pair:

(di ) = final − midterm

= (3.13, 5.07, 3.05, 8.74, 2.65, 11.91, 11.72, 0.71, 22.12, 2.03,
1.43, 10.00, −10.61, 7.79, 13.38, 2.24, 3.49, 0.43, 1.86, −12.50) .

1
 The mean of di is d¯ = 4.432. The sample variance based on di is
given by
1 X ¯ 2 = 60.00 .
s2 = (di − d)
20 − 1
The paired t-statistic has value

d¯
Tobs = = 2.559 .
(s2 /20)0.5
The reference distribution is t with 19 degrees of freedom. We
compute the p-value as

p = 2 × (1 − pt(2.559, 19)) = 0.019

for the two-sided alternative. Because 0.019 < 0.05, we reject the
null hypothesis that students perform similarly on the midterm
and the final (δ = 0) in favour of the claim that performances are
different at the conventional level 0.05.
(b) [4] If one mistakes the problem for a two-sample problem, what
would be the conclusion based on a two-sample t-test? (Perform
the standard t-test.)
Answer. The hypotheses are

H0 : µm = µf , H1 : µm ̸= µf .

In this case, the pooled variance estimator is

var(midterm) + var(final)
s2pool = = 125.12 .
2
The test statistic has observed value (in conventional notation)
ȳ2 − ȳ1
Tobs = q
1/20 + 1/20)s2pool
77.686 − 73.254
= q = 1.253 .
125.12
10

2
 The reference distribution is t with n1 + n2 − 2 = 38 and so the
p-value would be

p = 2 ∗ (1 − pt(1.253, 38)) = 0.218 .

We do not reject the null hypothesis that students perform sim-

ilarly on the midterm and the final (δ = 0) at the conventional
level 0.05.
(c) [4] Conduct a randomization test for the above data.
Answer. The hypotheses are stated in the question. Recall
ȳ2 − ȳ1 = 4.432. For the two-sided alternative, we calculate the
proportion the times when |ȳ2 − ȳ1 | < |ȳ2∗ − ȳ1∗ | where ∗ indicates a
hypothetical sample obtained by doing random flips for each pair
independently. The following code computes this proportion.

D.obs = mean(final - midterm)

NN = 50000
DD= rep(0, NN)
for (i in 1:NN) {
ind = 2*rbinom(20, 1, 0.5) - 1
DD[i] = sum(midterm*ind - final*ind) / 20
}
pp = 2*(sum(DD>D.obs) + 0.5*sum(DD==D.obs)) / NN

The p-value is found to be 0.01764. We reject the null hypothesis

that students perform similarly on the midterm and the final (δ =
0) in favour of the claim that performances are different at the
conventional level 0.05.
Notice that this p-value is close to the one obtained from the
paired t-test.
(d) [4] Suppose one wishes to conduct an independent study on whether
students tend to perform better in the final exam by an average of
3 marks or higher. The target is to obtain a significant outcome
at the 5% level with probability 0.8. Based on the above data

3
(assumed to be useful), how many pairs of observations should
she collect for this separate study?
Remark: independent thinking is needed as nothing discussed in
class can be directly applied.
Answer. We do a sample size calculation based on the paired
t-test. Note that other approaches are acceptable as long as they
are reasonable given the context and carried out correctly.
Let δ = µfinal − µmidterm . The null hypothesis is that H0 : δ ≤ 3
and the alternative is that H1 : δ > 3. Let n be the number of
pairs inPthe new study. Taking d¯new = ȳnew:final − ȳnew:midterm and
n
(di −d¯new )2
s2new = i=1 n−1 as defined in (a), the test statistic

d¯new − 3
t0 =
(s2new /n)0.5
has a t-distribution with n − 1 degrees of freedom under the as-
sumption δ = 3. The upper α = 0.05 quantile of this distribution
is given by qn = qt(1 − α, n − 1).
To do a power calculation, we use the current data to obtain an
effect size δ = d¯ = 4.432. Under the alternative δ = 4.432, the
statistic
d¯new − 4.432 1.132
t1 = = t0 −
(s2new /n)0.5 (s2new /n)0.5
has a t-distribution with n − 1 degrees of freedom. This implies
that we can compute the power for a given n as

1 − β = P (t0 > qn | δ = 4.432)

 
1.132 1.132
= P t0 − 2 > qn − 2 | δ = 4.432
(snew /n)0.5 (snew /n)0.5
 
1.132
= P t1 > qn − 2 | δ = 4.432 .
(snew /n)0.5

Note that s2new is also a random variable that makes computing

the probability inconvenient, and so for simplicity, we estimate it

4
 using s2 = 60 from the current data. To get 1 − β ≥ 0.8, we find
that we need at least n = 291 pairs in the new study.
R code:
nn = 10:300
alpha = 0.05
svar = var(final-midterm)
pows = pt(qt(1-alpha,nn-1)-1.132/sqrt(svar/nn), nn-1, lower.tail=F)
nn[which(pows > 0.8)[1]]

(e) [4] Does a student have a higher or lower probability of doing

better on the final compared to the midterm? Perform a test on
this dataset to answer this question.
Hint: the test you should use is not covered in STAT 404 (though
likely in previous statistics courses). Think of a common distribu-
tion that has probability as a parameter. The test is directly based
on this distribution.
Answer. The number of students who did better in the final is
a good metric of whether or not students perform similarly on
the final and midterm. If we denote this random variable as X,
then it has a binomial distribution with parameter n = 20 and
probability of success θ = 0.5 under the null assumption. That
is, we may use Binom(20, 0.5) as our reference distribution. The
alternative hypothesis is that θ ̸= 0.5.
In this context, we have Xobs = 18. The values 0, 1, 19, 20 would be
considered as more extreme observations in addition to the equally
extreme observations 2, 18. The p-value could be computed as

p = P (X ∈ {0, 1, 19, 20}) + 0.5P (X ∈ {2, 18}) = 0.00022 .

If one does not apply continuity correction, they would get

p = P (X ∈ {0, 1, 19, 20}) + P (X ∈ {2, 18}) = 0.00040 .

5
 Both are acceptable. The outcome is highly significant, and we
reject the null hypothesis that students perform similarly on the
midterm and the final at the conventional level 0.05.
Note: other tests are acceptable as long as they are reasonable
given the context and carried out correctly.

2. [20] Four students conducted an experiment on paper helicopters. The

response is the time it takes for a helicopter to touch the ground after
being dropped from 2 meters above ground. Four helicopter designs
are implemented. We use the data as if it were collected via a complete
randomized block design. The data for these four students are given
as yy1, yy2, yy3, yy4 below. Each column corresponds to a helicopter
design (a treatment).

yy1 = c(1.56, 1.62, 2.14, 1.30)

yy2 = c(1.53, 1.75, 2.02, 1.41)
yy3 = c(1.58, 1.80, 1.97, 1.36)
yy4 = c(1.60, 1.81, 1.93, 1.45)

This problem asks you to go over all routine data analysis for the
complete randomized design and a bit more.

(a) [6] Construct the analysis of variance table.

Answer. We compute the sum of squares of various effects.
First, the grand mean is found to be η̂ = ȳ.. = 1.676875.
The blocking effects of student volunteers are estimated to be

ȳi· − ȳ·· = (−0.021875, 0.000625, 0.000625, 0.020625).

Their SS is computed as
4
X
SSb = 4 (ȳi· − ȳ·· )2 = 0.00361875.
i=1

The treatment effects are estimated to be

ȳ·j − ȳ.. = (−0.109375, 0.068125, 0.338125, −0.296875)

6
 and the treatment SS is computed as
4
X
SStrt =4 (ȳ·j − ȳ·· )2 = 0.8762688.
j=1

The residuals have SS given by

4 X
X 4
SSerr = (yij − ȳi· − ȳ·j + ȳ·· )2 = 0.05945625.
i=1 j=1

The total SS is given by

4 X
X 4
SStot = (yij − ȳ·· )2 = 0.9393438
i=1 j=1

The MSS are obtained by dividing the SS by their corresponding

DF. The F-statistic is obtained by dividing the treatment MSS by
the error MSS. Hence, the ANOVA table is given by

Source DF SS MSS F
Volunteer 3 0.0036 0.0012
Treatment 3 0.8763 0.2921 44.21
Error 9 0.0594 0.0066
Total 15 0.9393

(b) [2] Test the hypothesis for H0 : treatment effects are the same.
Answer. The F-statistic is found in the ANOVA table. The
p-value is computed as

p = 1 − pf(44.21, 3, 9) = 1.035 × 10−5 .

The null hypothesis of no difference between treatment effects is

rejected at the 5% level.

7
(c) [8] Regardless of the outcome of (b), use Tukey’s method to con-
struct simultaneous confidence intervals for differences in treat-
ment means.
Answer. The pairwise treatment differences for (1, 2), (1, 3),
(1, 4), (2, 3), (2, 4), (3, 4) are estimated as

τ̂i − τ̂j = (−0.1775, −0.4475, 0.1875, −0.2700, 0.3650, 0.6350) .

The 95% Tukey quantile is given by

qtukey(0.95, 4, 9) = 4.41489 .

The width of the interval is given by

s 
qtukey(0.95, 4, 9) 1 1
√ + MSS(error) = 0.1794186 .
2 4 4

Hence, the simultaneous Tukey 95% CI for the difference between

treatments (1, 2), (1, 3), (1, 4), (2, 3), (2, 4), (3, 4) are given by

low -0.357 -0.627 0.008 -0.449 0.186 0.456

upp 0.002 -0.268 0.367 -0.091 0.544 0.814

All the differences except between treatments (1, 2) are found to

be significant.
R code:
eff.diff = c(ybar.trt[1]-ybar.trt[-1],
ybar.trt[2]-ybar.trt[c(3,4)],
ybar.trt[3]-ybar.trt[4])
ci.wdth = qtukey(0.95, 4, 9)*((1/4+1/4)*MSS.err/2)^.5
low = eff.diff - ci.wdth
upp = eff.diff + ci.wdth

(d) [4] Construct a 95% two-sided CI for τ1 + τ3 − 2τ2 using the rec-
ommended universal recipe.

8
 Answer. We estimate θ = τ1 + τ3 − 2τ2 by
θ̂ = (ȳ·1 − ȳ·· ) + (ȳ·3 − ȳ·· ) − 2(ȳ·2 − ȳ·· )
= ȳ·1 + ȳ·3 − 2ȳ·2 = 0.0925 .
When regarded as a random variable (rather than an observed
value), θ̂ is the linear combination of 3 independent means where
each is a mean of 4 observations. Hence, we have
3
Var(θ̂) = (1/4 + 1/4 + 1)σ 2 = σ 2 .
2
We naturally estimate it by
d θ̂) = 3 MSS(error) = 0.0099
Var(
2
which has 9 degrees of freedom. Note qt(0.975, 9) = 2.262157.
Hence, a 95% CI for θ = τ1 + τ3 − 2τ2 is given by
√
0.0925 ± 2.262157 0.0099 = (−0.133, 0.318) .

3. [20] Animals were randomly allocated to 12 groups of 4. Each group

was given one of 3 poisons and one of 4 treatments. The survival times
of the animals are given in the following table.
treatment
poison A B C D
I 0.31 0.82 0.43 0.45
0.45 1.10 0.45 0.71
0.46 0.88 0.63 0.66
0.43 0.72 0.76 0.62
II 0.36 0.92 0.44 0.56
0.29 0.61 0.35 1.02
0.40 0.49 0.31 0.71
0.23 1.24 0.40 0.38
III 0.22 0.30 0.23 0.30
0.21 0.37 0.25 0.36
0.18 0.38 0.24 0.31
0.23 0.29 0.22 0.33

9
(Source: Box, Hunter, and Hunter, altered slightly).
The following code might be useful.

n = 4
k1 = 4
k2 = 3
dat = data.frame(times = c(0.31,0.82,0.43,0.45,
0.45,1.10,0.45,0.71,
0.46,0.88,0.63,0.66,
0.43,0.72,0.76,0.62,
0.36,0.92,0.44,0.56,
0.29,0.61,0.35,1.02,
0.40,0.49,0.31,0.71,
0.23,1.24,0.40,0.38,
0.22,0.30,0.23,0.30,
0.21,0.37,0.25,0.36,
0.18,0.38,0.24,0.31,
0.23,0.29,0.22,0.33),
group = rep(c("A","B","C","D"), n*k2),
poison = rep(c("I","II","III"), each=n*k1))

(a) [8] Construct the analysis of variance table for this two-way layout
design.
Answer. The calculations for the SS corresponding to the main
effects are similar to that described in Q(2a). The interaction SS
is given by
3 X
X 4
SSint =4 (ȳij. − ȳi.. − ȳ.j. + ȳ... )2 .
i=1 j=1

The error SS is found by subtracting all of the other SS from the

total SS.
The ANOVA table is given by

10
 Source DF SS MSS F
Treatment 3 0.9212 0.3071 13.80
Poison 2 1.033 0.5165 23.22
Interaction 6 0.2501 0.0417 1.874
Error 36 0.8007 0.0222
Total 47 2.0014

R code:
I = 4
J = 3
n = 4

yy1 = matrix(c(0.31,0.82,0.43,0.45,
0.45,1.10,0.45,0.71,
0.46,0.88,0.63,0.66,
0.43,0.72,0.76,0.62), 4, 4, byrow=T)
yy2 = matrix(c(0.36,0.92,0.44,0.56,
0.29,0.61,0.35,1.02,
0.40,0.49,0.31,0.71,
0.23,1.24,0.40,0.38), 4, 4, byrow=T)
yy3 = matrix(c(0.22,0.30,0.23,0.30,
0.21,0.37,0.25,0.36,
0.18,0.38,0.24,0.31,
0.23,0.29,0.22,0.33), 4, 4, byrow=T)

mean.trt = colMeans(rbind(yy1,yy2,yy3))
mean.poison = c(mean(yy1), mean(yy2), mean(yy3))
mean.int = rbind(colMeans(yy1),colMeans(yy2),colMeans(yy3))
ybar = mean(mean.trt)

# Compute SS
SS.trt = n*J*sum((mean.trt-ybar)^2)
SS.poison = n*I*sum((mean.poison-ybar)^2)
SS.int = n*sum((mean.int
-matrix(rep(mean.poison,I),J,I,byrow=F)
-matrix(rep(mean.trt,J),J,I,byrow=T)

11
 +ybar)^2)
SS.total = sum((yy1-ybar)^2) + sum((yy2-ybar)^2) + sum((yy3-ybar)^2)
SS.err = SS.total - SS.poison - SS.trt - SS.int

# Compute MSS
MS.trt = SS.trt / (I-1)
MS.poison = SS.poison / (J-1)
MS.int = SS.int / ((I-1)*(J-1))
MS.err = SS.err / (I*J*(n-1))

# Compute F-statistic
F.trt = MS.trt / MS.err
F.poison = MS.poison / MS.err
F.int = MS.int / MS.err

(b) [4] Test the hypotheses of whether the effects of poison and treat-
ment are significant.
Answer. For each of the factors, we test the hypotheses that the
main effects are equal (H0 ) against the alternative that at least
two of the levels have different effects.
The 95% quantiles of F3,36 and F2,36 are 2.866266 and 3.259446,
respectively.
Using these reference distributions and the observed F-statistics
obtained in (a), we reject H0 for both treatment and poison and
conclude that both have significant effects.
R code:
qf(0.95, 3, 36)
qf(0.95, 2, 36)

(c) [4] Test the hypothesis of whether the interaction effect is signifi-
cant.
Answer. For the interaction, we test the hypotheses that the
interaction effects are equal (H0 ) against the alternative that the
effects are not all equal.

12
 The 95% quantile of F6,36 is 2.363751.
Using this reference distribution and the observed F-statistic ob-
tained in (a), we do not reject H0 and conclude that the interaction
between treatment and poison is not significantly different from
0.
R code:
qf(0.95, 6, 36)

(d) [4] Predict the survival time of an animal administered poison II

and treatment C. Estimate the variance of the prediction.
Answer. In (c), we concluded that the interaction between poison
and treatment is insignificant. Hence, our predicted survival time
of an animal administered poison II (α2 ) and treatment C (β3 ) is
given by
ŷ = η̂ + α̂2 + β̂3 = 0.4575 units .
Under standard assumptions of the model (independent units),
the variance is given by

Var(ŷ) = Var(η̂ + α̂2 + β̂3 )

= Var(ȳ... + (ȳ2.. − ȳ... ) + (ȳ.3. − ȳ... ))
= Var(ȳ2.. + ȳ.3. − ȳ... )
= Var(ȳ2.. ) + Var(ȳ.3. ) + Var(ȳ... )
+ 2Cov(ȳ2.. , ȳ.3. ) − 2Cov(ȳ2.. , ȳ... ) − 2Cov(ȳ.3. , ȳ... )
= Var(ȳ2.. ) + Var(ȳ.3. ) + Var(ȳ... )
2×4 2 × 16 2 × 12
+ Var(ȳ231 ) − Var(y211 ) − Var(y131 )
12 × 16 16 × 48 12 × 48
σ2 σ2 σ2 σ2 σ2 σ2
= + + + − −
16 12 48 24 24 24
σ2
= .
8
Estimating σ 2 by MSSerr , we get Var(ŷ)
d = 0.0028.

13