Mock exam
50 minutes · no calculator, no tables · your one sheet of notes · answer in full sentences
Question 1
A coffee chain wants to estimate the mean amount of caffeine in its medium cold brew. A random sample of 64 drinks gives a mean of 200 mg and a standard deviation of 16 mg. A 95% confidence interval for the mean is (196, 204) mg.
(a) Interpret this interval.
Answer
We are 95% confident that the true mean caffeine content of all medium cold brews this chain sells is between 196 mg and 204 mg.
(b) Is the interval statistically valid? Explain.
Answer
Yes. The 64 drinks are a random sample, and 64 is at least 30, so by the Central Limit Theorem the sample mean is approximately normal whatever the shape of individual caffeine amounts.
(c) True or false, and explain: 95% of medium cold brews contain between 196 and 204 mg of caffeine.
Answer
False. The interval estimates the mean of all drinks. It does not describe individual drinks. Single drinks vary with a standard deviation of about 16 mg, so many fall outside a range only 8 mg wide.
(d) The manager asks for a 99% interval "to be more precise". Without recalculating, explain what would happen.
Answer
The interval would be wider, so it would be less precise. The mean, standard deviation and sample size are unchanged, and the multiplier for 99% is larger than for 95%, so the margin of error grows. More confidence costs precision. To get a narrower interval at 95%, take a larger sample.
(e) What is the margin of error of the given interval? What sample size would cut it in half at the same confidence level?
Answer
Margin = (204 − 196)/2 = 4 mg. Halving the margin needs 4 times the sample size, so 4 × 64 = 256 drinks.
Question 2
A city plans a survey to estimate the proportion of residents who support a new bus route.
(a) With no earlier estimate, how many completed surveys are needed for a margin of error of 0.05 at 95% confidence? Use 1.96² ≈ 3.84.
Answer
Use p = 0.5. n = 3.84 × 0.25/0.05² = 0.96/0.0025 = 384. With the exact 3.8416 it is 384.2, so round up to 385.
(b) Why is p = 0.5 used?
Answer
The product p(1 − p) is largest at 0.5, where it is 0.25. That gives the largest sample size, so the margin of error is met whatever the true proportion is.
(c) Only 40% of residents contacted complete the survey. How many must be contacted?
Answer
385/0.40 = 962.5, so round up to 963.
(d) The survey is completed by 400 residents and 80 support the route. Let X be the number of supporters in a new sample of 400, assuming the true proportion is 0.20. State the distribution of X, its mean and its standard deviation.
Answer
X is binomial with n = 400 and p = 0.20. Mean = 80. Variance = 400 × 0.2 × 0.8 = 64, so the standard deviation is 8.
(e) Can a normal approximation be used for X? If so, write the z-value for the probability of more than 96 supporters, and estimate that probability.
Answer
Yes. The expected successes (80) and failures (320) are both at least 5.
“More than 96” leaves 96 out, so use 96.5. z = (96.5 − 80)/8 = 2.06. That is about 2 standard deviations above the mean, so by the empirical rule the probability is about 2.5% (a little less, since 2.06 is slightly past 2).
Question 3
An order passes through three independent stages. The stage times in minutes are normal with means 10, 20 and 30 and variances 4, 5 and 16.
(a) Describe the distribution of the total time for one order.
Answer
Normal, because it is a sum of independent normals. Mean = 10 + 20 + 30 = 60 minutes. Variance = 4 + 5 + 16 = 25, so the standard deviation is 5 minutes.
(b) About what is the probability that an order takes more than 70 minutes? Between 55 and 65 minutes?
Answer
70 is 2 standard deviations above the mean, so about 2.5%.
55 to 65 is within 1 standard deviation either side, so about 68%.
(c) Four orders are independent. Write an expression for the probability that exactly one takes more than 70 minutes.
Answer
The number of slow orders is binomial with n = 4 and p = 0.025. P(exactly 1) = C(4,1) × 0.025 × 0.975³ = 4 × 0.025 × 0.975³.
(d) Describe the distribution of the average time of 25 independent orders. About what is the probability the average exceeds 62 minutes?
Answer
Normal with mean 60 and variance 25/25 = 1, so the standard deviation is 1 minute. 62 is 2 standard deviations above the mean, so about 2.5%.
(e) If the stage times were not normal, could you still answer (a) and (d)? Explain.
Answer
Not (a). With only 3 stages the Central Limit Theorem does not apply, so the shape of the total would be unknown. The mean of 60 and variance of 25 would still be right.
Part (d) would be borderline. The Central Limit Theorem needs about 30 or more observations, and 25 orders is a little short, so the normal shape would be only a rough approximation.