Problems — Easy (1–20)
1. In a provincial town, previous surveys show 70% of households own a cellphone. A local
marketing student thinks the proportion may have changed. He interviews 200
households and 130 say they own a cellphone. At the 95% confidence level (p-value
method), test whether the true cellphone ownership proportion differs from 70%.
2. A city’s public health office reports the average adult weight is 168 lb. A nutritionist
suspects that community programs altered this. She measures 36 random adults and
finds a mean of 169.5 lb with sample standard deviation 3.9 lb. Use the p-value method
at 95% confidence to test whether the mean differs from 168 lb.
3. A car manufacturer gives a 5-year warranty on engines. An engineer suspects failures
occur sooner than 5 years. She samples 40 returned cars and finds average lifetime until
failure 4.8 years with s = 0.50 years. At α = 0.02, test if mean lifetime is less than 5 years
(p-value method).
4. The local library claims that the average time patrons spend reading per visit is 45
minutes. A program evaluation sampled 25 patrons and found an average of 48 minutes
with s = 6 minutes. Test at α = 0.05 (p-value) whether reading time increased.
5. A grocery chain says 80% of its customers use the loyalty card. A regional manager thinks
usage is lower. She samples 150 shoppers and finds 110 used the card. At 5%
significance (p-value), test whether the cashier claim is too optimistic.
6. A teacher says her students average 72% on a standardized quiz. After a new teaching
method, 30 students have mean 74.5% and s = 5%. At α = 0.05, test if the new method
improved scores (p-value).
7. A software company advertises that average install time is 10 minutes. A tester samples
20 installs (s = 2 minutes) and sees mean 11.2 minutes. At α = 0.05, test whether install
time is longer than claimed (p-value).
8. A local clinic claims that 12% of visitors require a follow-up within a week. A nurse
randomly samples 200 visits and finds 30 follow-ups. At 5% significance, test whether
the follow-up rate differs from 12% (p-value).
9. A bakery claims loaf weight averages 500 g. A quality inspector samples 16 loaves, mean
= 495 g, s = 8 g. At α = 0.05, test if loaves weigh less than claimed (p-value).
10. A university claims average library use is 7 hours per week. A student survey of 36
students yields mean 6.7 hours, s = 1.2 hours. At α = 0.05, test whether weekly library
use decreased (p-value).
11. A small apparel factory states that 95% of items pass quality checks. A new inspector
samples 100 items and finds 90 passing. At the 5% level, test whether the true pass rate
is lower (p-value).
12. A gym advertises average class attendance of 25 people. After a schedule change, a
manager samples 20 classes and finds mean attendance 23 with s = 3. At α = 0.05, test if
attendance decreased (p-value).
13. A traffic analyst claims the mean commute time along a route is 35 minutes. A sample of
30 commuters records mean 36.8 minutes and s = 4.5 minutes. At 5% significance, test
whether commute time increased (p-value).
14. A pharmacist claims average pill weight is 250 mg. A batch test of 10 pills shows mean
249 mg with s = 1 mg. At α = 0.05, test if pills are underweight (p-value).
15. A fast-food chain reports average service time 3.5 minutes. A mystery shopper times 40
customers and finds mean 3.7 min, s = 0.9 min. Test at α = 0.05 if service time increased
(p-value).
16. A company runs a telemarketing campaign that historically converts at 8%. After a script
change, 150 calls get 15 conversions. At 5% significance, test whether conversion rate
changed (p-value).
17. A local pool reports average water temperature 78°F. After maintenance, 20 readings
average 77.2°F with s = 0.8°F. At α = 0.05, test whether temperature decreased (p-value).
18. An online game reports average daily playtime 120 minutes. A patch is released and a
sample of 30 users shows mean 126 min with s = 20. At α = 0.05, test if playtime
increased (p-value).
19. A driver school claims 90% of students pass the road test on the first attempt. In a class
of 80 students, 68 pass first time. At 5% level, test if pass rate equals claimed value (p-
value).
20. A city claims average daily public transit riders are 10,000. A researcher samples 25 days
and gets mean 9,700 with s = 400. At α = 0.05, test whether ridership decreased (p-
value).
Problems — Medium (21–40) — trickier / more detailed
21. A beverage plant fills bottles claiming 500 ml on average. A complaint suggests
underfilling. A worker samples 60 bottles and obtains mean 497 ml and variance 9 ml².
Using the p-value method at α = 0.05, test whether the machine is underfilling.
22. A municipal survey reports average household electricity bill is $120. After a rate change,
a sample of 24 households shows mean $127 with s = $15. At 95% confidence (p-value
method), test if the mean bill increased.
23. A biotech firm claims a new test's accuracy is 98%. An independent lab tests 250
specimens and gets 242 accurate results. At α = 0.05, use the p-value method to assess
the claim.
24. A school district says that on average students spend 2 hours per night on homework. A
pilot after a curriculum change finds for 22 students mean 2.3 hours, s = 0.45 hours. At α
= 0.05, test if homework time increased using p-values.
25. An engine part is claimed to last 10,000 miles on average. Field testing on 18 parts yields
mean life 9,650 miles with s = 420 miles. At α = 0.05, test whether the mean life is less
than claimed (p-value).
26. A streaming platform reports 40% of subscribers watch a flagship show weekly. A
content analyst samples 150 subscribers and finds 68 watched. At 5% significance, test if
this proportion differs (p-value method).
27. A hospital claims average patient wait is 20 minutes. Patients complain after triage
changes. Sample of 30 wait times: mean 22.5 min, s = 6. At α = 0.05, test if wait
increased (p-value).
28. A coffee roastery reports average roast time 18 minutes with σ = 1.2 minutes (pop σ
known). A batch sample of 25 roasts yields mean 18.4 min. At α = 0.05 using p-value
method, test whether roast time increased.
29. A factory accuracy test says 2% of units are defective. A sample of 300 units shows 11
defective. At α = 0.05, test whether defect proportion exceeds 2% (p-value).
30. A manager believes a recent training raised average shop productivity from 75 units/day.
A sample of 20 workers after training reports mean 79 units/day with s = 6. At α = 0.05,
use p-value to test for increase.
31. A smartphone maker claims average battery life 24 hr (σ known = 2 hr). A tech reviewer
tests 36 units, mean 23.2 hr. At α = 0.05, use p-value to test if battery life is less.
32. An airline states on-time departures are 92%. A traveler group samples 200 flights and
finds 170 on-time. At α = 0.05, use p-value to check the claim.
33. A dietician claims average daily sodium intake is 2,300 mg. A community sample of 28
records mean 2,470 mg, s = 300 mg. At α = 0.05, test whether intake is higher (p-value).
34. A packaging machine target weight is 250 g; the machine is supposed to have σ = 3 g. A
sample of 40 packages yields mean 248.7 g. Using p-value at α = 0.05, test for underfill.
35. A university claims 60% of graduates find jobs within 6 months. A study of 120 grads
shows 68 found jobs in that time. At α = 0.05, test if the rate differs (p-value).
36. A farm cooperative claims average crop yield 1.8 tons/ha. A sample of 15 plots under a
new fertilizer yields mean 1.95 tons/ha, s = 0.25. At α = 0.05, test if fertilizer increases
yield (p-value).
37. A car rental company states average fuel economy is 30 mpg. After a software update, a
sample of 22 cars shows mean 31.6 mpg with s = 2.8 mpg. At α = 0.05, test for an
increase (p-value).
38. A medical device passes quality if ≤1% malfunction. A test of 500 devices shows 7
malfunctions. At α = 0.05, test whether malfunction rate exceeds 1% (p-value).
39. A bank claims average customer waiting time 8 minutes. An auditor samples 30
customers and finds mean 9.1 min, s = 2. At α = 0.05, test if waiting time increased (p-
value).
40. A traffic study claims the average red-light running incidents per week at an intersection
is 3 with σ = 1.2. After installing sensors, a 36-week sample shows mean 3.5. At α = 0.05,
use p-value to test whether incidents increased.
Problems — Hard (41–60) — tricky & ambiguous; decide tail carefully
41. Regulators claim that at most 5% of facilities in a region
violate a safety code. A surprise audit of 80 facilities finds
7 violations. With α = 0.01 and p-value method, test
whether the regulator’s assertion still holds; state the
appropriate alternative.
42. A clinical benchmark states mean systolic blood pressure
for a healthy cohort is 120 mmHg. A new drug trial (n =
12) shows mean 128 mmHg with s = 9 mmHg. At α = 0.01
and using p-value method, evaluate whether the drug
raises blood pressure.
43. A stadium reports 65% average seat occupancy for
regular games. A recent promotional event sampled 150
ticket scans and recorded 95 occupied seats. At α = 0.01,
use p-value to test if occupancy rate deviates from 65%
(decide one/two-tailed).
44. A precision tool maker claims mean dimension 10.00
mm. A prototype batch of 10 parts has mean 9.92 mm, s
= 0.12 mm. At α = 0.01, use p-value to test if dimensions
have shifted (careful with tail).
45. A national exam historically has a pass rate of 78%. After
a change in test format, 500 candidates produce 365
passes. At α = 0.01, use the p-value method to see if pass
rate changed.
46. A biotech company claims an assay detects disease with
95% specificity. An independent lab runs 200 healthy
samples; 8 are falsely flagged. At α = 0.01 use p-value to
evaluate the claim.
47. A research lab believes a new protocol will decrease
reaction time from 50 seconds. A pilot of 9 trials yields
mean 46.5 s, s = 4.2 s. Using α = 0.01 and p-value
method, test whether reaction time decreased.
48. A city claims that only 10% of households are energy
inefficient. A comprehensive survey of 120 homes shows
18 inefficient. At α = 0.01, test the city’s claim with p-
value method.
49. A drug’s adverse event rate is reported at 2%. In a clinical
monitoring trial of 300 patients, 12 experienced the
event. At α = 0.01, test whether adverse rate has
increased.
50. A tech firm sets mean server response time 200 ms. After
an update a small sample of 12 tests shows mean 215 ms
with s = 15 ms. At α = 0.01, evaluate if response time
increased (p-value).
51. An educational program targets average improvement of
6 points. A small trial of 10 students yields mean
improvement 8.2 with s = 3.5. Use α = 0.01 to test
whether program achieves more than 6 (p-value).
52. A quality control standard allows up to 4% defective
items. A random batch of 75 items contains 5 defective.
At α = 0.01 and p-value method, test whether defect
level exceeds standard.
53. A hospital expects average length of stay 5 days. After a
clinical protocol change, a pilot of 11 patients shows
mean 4.2 days, s = 1.1 days. At α = 0.01, test if length of
stay decreased (p-value).
54. A software QA benchmark shows 0.5% crash rate. In a
stress test of 1000 runs, 10 crashes occur. At α = 0.01,
test whether crash rate exceeds benchmark (p-value).
55. A precision clock has mean daily drift 0.2 sec with σ =
0.05 sec. A new design tested for 12 days gives mean
drift 0.25 sec. At α = 0.01, test whether drift increased (p-
value, σ known).
56. A clinical recommendation cites 90% adherence. A
hospital audit of 30 records shows 25 adhered. Using α =
0.01 and p-value method, test whether adherence is
lower than recommended.
57. A farmer expects mean yield 3.5 tons/ha. A trial of 14
plots with a new seed shows mean 3.2, s = 0.4. At α =
0.01, test if yield decreased (p-value).
58. A blood screening test is claimed to have mean
processing time 45 minutes. An independent lab (n = 16)
finds mean 47.8 minutes, s = 3.7 min. At α = 0.01,
evaluate increase (p-value).
59. A manufacturer advertises average battery capacity 3000
mAh. In small sample of 10 units mean capacity is 2910
mAh with s = 120. At α = 0.01, test whether capacity is
less (p-value).
60. A civic study claims 30% of residents cycle to work. An
advocacy group samples 120 residents and finds 34 cycle.
At α = 0.01, use p-value to test whether cycling rate is
greater than 30% (carefully decide one/two-tailed).
Answers — full 6-step p-value method
Below are answers for all 60 problems. Each problem’s answer uses the p-value method and
lists:
1. H₀, 2) Hₐ, 3) α and test type (z or t or proportion z), 4) decision rule (p-value), 5) test
statistic and p-value, 6) decision & conclusion (practical interpretation).
Note: For proportion tests we use the large-sample z-test for proportions:
( z = \dfrac{\hat p - p_0}{\sqrt{p_0(1-p_0)/n}} ).
For mean tests with known σ use z; with unknown σ use t with df = n−1. If sample is large (e.g.
n≥30) using z is acceptable. I compute p-values accordingly.
EASY answers (1–20)
1. (Cellphone ownership, n=200, x=130)
1. H₀: p = 0.70
2. Hₐ: p ≠ 0.70 (two-sided)
3. α = 0.05, z-test for proportion.
4. Decision: reject H₀ if p-value < 0.05.
5. p̂ = 130/200 = 0.65. SE = sqrt(0.70.3/200) ≈ 0.0324. z = (0.65−0.7)/0.0324 ≈ −1.543. p-
value(two) ≈ 2Φ(−1.543) ≈ 0.123.
6. p-value 0.123 > 0.05 → fail to reject H₀. Conclusion: Insufficient evidence that cellphone
ownership differs from 70%.
2. (Weight, n=36, x̄ =169.5, s=3.9, μ₀=168)
1. H₀: μ = 168
2. Hₐ: μ ≠ 168 (two-sided)
3. α = 0.05, t-test (σ unknown, n=36 but t used). df = 35.
4. Reject if p-value < 0.05.
5. SE = 3.9/√36 = 0.65. t = (169.5−168)/0.65 ≈ 2.308. p-value(two) ≈ 2*P(T>2.308, df=35) ≈
~0.027 (approx).
6. p-value ~0.027 < 0.05 → reject H₀. Conclusion: Evidence that mean weight differs (it
appears increased).
3. (Warranty, n=40, x̄ =4.8, s=0.50, μ₀=5, α=0.02, left tail)
1. H₀: μ = 5
2. Hₐ: μ < 5
3. α = 0.02, t-test (σ unknown), df=39.
4. Reject if p-value < 0.02.
5. SE = 0.50/√40 ≈ 0.07906. t = (4.8−5)/0.07906 ≈ −2.5298. p-value = P(T < −2.53) ≈ 0.007
(approx).
6. p-value ≈ 0.007 < 0.02 → reject H₀. Conclusion: Evidence that mean life is less than 5
years; warranty may need revision.
4. (Library reading, n=25, x̄=48, s=6, μ₀=45, α=0.05)
1. H₀: μ = 45
2. Hₐ: μ > 45
3. α = 0.05, t-test, df=24.
4. Reject if p-value < 0.05.
5. SE = 6/√25 = 1.2. t = (48−45)/1.2 = 2.5. p-value = P(T>2.5) ≈ 0.009–0.01.
6. p-value < 0.05 → reject H₀. Conclusion: Reading time increased.
5. (Loyalty card, n=150, x=110)
1. H₀: p = 0.80
2. Hₐ: p < 0.80 (manager thinks lower)
3. α = 0.05, z-test for proportion.
4. Reject if p-value < 0.05.
5. p̂ = 110/150 ≈ 0.7333. SE = sqrt(0.8*0.2/150) ≈ 0.03266. z = (0.7333−0.8)/0.03266 ≈
−2.044. p-value = P(Z < −2.044) ≈ 0.0205.
6. p-value ≈ 0.0205 < 0.05 → reject H₀. Conclusion: Evidence loyalty usage is lower than
80%.
6. (Quiz scores, n=30, x̄ =74.5, s=5, μ₀=72)
1. H₀: μ =72
2. Ha: μ > 72
3. α=0.05, t-test df=29.
4. Reject if p-value < 0.05.
5. SE = 5/√30 ≈ 0.9129. t = (74.5−72)/0.9129 ≈ 2.742. p-value = P(T>2.742) ≈ 0.005–0.01.
6. p-value < 0.05 → reject H₀. Conclusion: Evidence the method improved scores.
7. (Install time, n=20, x̄ =11.2, s=2, μ₀=10)
1. H₀: μ = 10
2. Ha: μ > 10
3. α=0.05, t-test df=19.
4. Reject if p-value < 0.05.
5. SE=2/√20=0.4472. t=(11.2−10)/0.4472 ≈ 2.687. p-value ≈ 0.007–0.01.
6. p-value < 0.05 → reject H₀. Conclusion: Install time longer than claimed.
8. (Follow-up rate, n=200, x=30)
1. H₀: p = 0.12
2. Ha: p ≠ 0.12
3. α=0.05, z-test for proportion.
4. Reject if p-value < 0.05.
5. p̂ =30/200=0.15. SE=sqrt(0.12*0.88/200)=0.0230. z=(0.15−0.12)/0.0230≈1.304. two-
tailed p ≈ 0.192.
6. p-value>0.05 → fail to reject H₀. Conclusion: Not enough evidence to say follow-up rate
differs from 12%.
9. (Loaf weight, n=16, x̄ =495, s=8, μ₀=500)
1. H₀: μ = 500
2. Ha: μ < 500
3. α=0.05, t-test df=15.
4. Reject if p-value < 0.05.
5. SE = 8/4 = 2.0. t=(495−500)/2 = −2.5. p-value = P(T < −2.5) ≈ 0.011–0.02.
6. p-value < 0.05 → reject H₀. Conclusion: Loaves underweight.
10. (Library hours, n=36, x̄ =6.7, s=1.2, μ₀=7)
1. H₀: μ = 7
2. Ha: μ < 7
3. α=0.05, t-test df=35. (n large enough but sigma unknown)
4. Reject if p-value < 0.05.
5. SE = 1.2/6 = 0.2. t = (6.7−7)/0.2 = −1.5. p-value ≈ 0.071 (one-tailed).
6. p-value > 0.05 → fail to reject H₀. Conclusion: Insufficient evidence that library use
decreased.
11. (Apparel QC, n=100, x=90 pass)
1. H₀: p = 0.95
2. Ha: p < 0.95
3. α=0.05, z-test proportion.
4. Reject if p-value<0.05.
5. p̂ =0.9. SE = sqrt(0.95*0.05/100)=0.02179. z=(0.9−0.95)/0.02179≈−2.293. p-value≈0.011.
6. p-value<0.05 → reject H₀. Conclusion: Pass rate lower than 95%.
12. (Gym attendance, n=20, x̄ =23, s=3, μ₀=25)
1. H₀: μ = 25
2. Ha: μ < 25
3. α=0.05, t-test df=19.
4. Reject if p-value < 0.05.
5. SE = 3/√20 ≈ 0.6708. t = (23−25)/0.6708 ≈ −2.981. p-value ≈ 0.004.
6. reject H₀. Conclusion: Attendance decreased.
13. (Commute, n=30, x̄ =36.8, s=4.5, μ₀=35)
1. H₀: μ = 35
2. Ha: μ > 35
3. α=0.05, t-test df=29.
4. Reject if p-value < 0.05.
5. SE = 4.5/√30 ≈ 0.8216. t = 1.8/0.8216 ≈ 2.19. p-value ≈ 0.018 (one-tailed) or ~0.036 two-
tailed.
6. p-value one-tailed <0.05 → reject H₀. Conclusion: Commute increased.
14. (Pill weight, n=10, x̄ =249, s=1, μ₀=250)
1. H₀: μ = 250
2. Ha: μ < 250
3. α=0.05, t-test df=9.
4. Reject if p-value < 0.05.
5. SE = 1/√10 ≈ 0.3162. t = (249−250)/0.3162 ≈ −3.162. p-value ≈ 0.006.
6. reject H₀. Conclusion: Pills underweight.
15. (Service time, n=40, x̄ =3.7, s=0.9, μ₀=3.5)
1. H₀: μ = 3.5
2. Ha: μ > 3.5
3. α=0.05, t-test df=39.
4. Reject if p-value < 0.05.
5. SE = 0.9/√40 ≈ 0.1421. t = 0.2/0.1421 ≈ 1.408. p-value ≈ 0.083.
6. p-value > 0.05 → fail to reject H₀. Conclusion: Not significant increase.
16. (Conversion, n=150, x=15)
1. H₀: p = historical (8%); since the question: test change (two-sided)
(Assume H₀: p = 0.08)
2. Ha: p ≠ 0.08
3. α=0.05, z-test.
4. Reject if p-value < 0.05.
5. p̂ =15/150=0.10. SE = sqrt(0.08*0.92/150)=0.02198. z=(0.10−0.08)/0.02198 ≈ 0.910. p-
value≈0.362.
6. fail to reject H₀. Conclusion: No evidence conversion changed.
17. (Pool temperature, n=20, x̄ =77.2, s=0.8, μ₀=78)
1. H₀: μ = 78
2. Ha: μ < 78
3. α=0.05, t-test df=19.
4. Reject if p-value < 0.05.
5. SE = 0.8/√20 ≈ 0.1789. t = (77.2−78)/0.1789 ≈ −4.47. p-value ≈ 0.0002.
6. reject H₀. Conclusion: Temperature decreased.
18. (Playtime, n=30, x̄ =126, s=20, μ₀=120)
1. H₀: μ = 120
2. Ha: μ > 120
3. α=0.05, t-test df=29.
4. Reject if p-value < 0.05.
5. SE = 20/√30 ≈ 3.6515. t = 6/3.6515 ≈ 1.643. p-value ≈ 0.055 (one-tailed).
6. p-value > 0.05 → fail to reject H₀ (marginal).
19. (Pass rate, n=80, x=68)
1. H₀: p = 0.90
2. Ha: p ≠ 0.90
3. α=0.05, z-test proportion.
4. Reject if p-value < 0.05.
5. p̂ =68/80=0.85. SE=sqrt(0.9*0.1/80)=0.03354. z=(0.85−0.9)/0.03354≈−1.49. two-tailed
p≈0.136.
6. fail to reject H₀. Conclusion: No evidence pass rate differs.
20. (Transit ridership, n=25, x̄=9700, s=400, μ₀=10000)
1. H₀: μ = 10000
2. Ha: μ < 10000
3. α=0.05, t-test df=24.
4. Reject if p-value < 0.05.
5. SE = 400/√25 = 80. t = (9700−10000)/80 = −300/80 = −3.75. p-value ≈ 0.0006.
6. reject H₀. Conclusion: Ridership decreased.
MEDIUM answers (21–40)
21. (Beverage fill, n=60, x̄ =497, variance 9 → s=3, μ₀=500, α=0.05, left)
1. H₀: μ = 500
2. Ha: μ < 500
3. α=0.05, t-test (s=3, n=60) — n large so t≈z possible; df=59.
4. Reject if p-value < 0.05.
5. SE = 3/√60 ≈ 0.3873. t = (497−500)/0.3873 ≈ −7.746. p-value ≈ ≈0 (very small)
6. reject H₀. Conclusion: Machine is underfilling.
22. (Electric bill, n=24, x̄=127, s=15, μ₀=120, α=0.05, right)
1. H₀: μ = 120
2. Ha: μ > 120
3. α=0.05, t-test df=23.
4. Reject if p < 0.05.
5. SE = 15/√24 ≈ 3.062. t = 7/3.062 ≈ 2.286. one-tailed p ≈ 0.016.
6. reject H₀. Conclusion: Average bill increased.
23. (Accuracy 98%, n=250, x=242)
1. H₀: p = 0.98
2. Ha: p ≠ 0.98
3. α=0.05, z-test proportion.
4. Reject if p-value < 0.05.
5. p̂ =242/250=0.968. SE = sqrt(0.98*0.02/250)=0.008912. z=(0.968−0.98)/0.008912 ≈
−1.347. two-tailed p ≈ 0.178.
6. fail to reject H₀. Conclusion: No evidence accuracy differs from 98%.
24. (Homework time, n=22, x̄ =2.3, s=0.45, μ₀=2, α=0.05, right)
1. H₀: μ = 2
2. Ha: μ > 2
3. α=0.05, t-test df=21.
4. Reject if p < 0.05.
5. SE = 0.45/√22 ≈ 0.0959. t=(2.3−2)/0.0959 ≈ 3.129. p ≈ 0.003–0.004.
6. reject H₀. Conclusion: Homework time increased.
25. (Engine life, n=18, x̄ =9650, s=420, μ₀=10000, α=0.05, left)
1. H₀: μ = 10000
2. Ha: μ < 10000
3. α=0.05, t-test df=17.
4. Reject if p < 0.05.
5. SE = 420/√18 ≈ 99.0. t = (9650−10000)/99 ≈ −3.535. p ≈ 0.0019.
6. reject H₀. Conclusion: Lifetime less than claimed.
26. (Show watchers, n=150, x=68)
1. H₀: p = 0.40
2. Ha: p ≠ 0.40
3. α=0.05, z-test proportion.
4. Reject if p < 0.05.
5. p̂ =68/150≈0.4533. SE = sqrt(0.4*0.6/150)=0.0400. z=(0.4533−0.4)/0.04≈1.333. two-
tailed p≈0.182.
6. fail to reject H₀. Conclusion: No evidence proportion changed.
27. (Wait time, n=30, x̄ =22.5, s=6, μ₀=20)
1. H₀: μ = 20
2. Ha: μ > 20
3. α=0.05, t-test df=29.
4. Reject if p < 0.05.
5. SE = 6/√30 ≈1.095. t = 2.5/1.095 ≈2.282. p ≈0.015.
6. reject H₀. Conclusion: Wait increased.
28. (Roast time σ known, n=25, x̄=18.4, σ=1.2, μ₀=18)
1. H₀: μ = 18
2. Ha: μ > 18
3. α=0.05, z-test (σ known).
4. Reject if p < 0.05.
5. SE = 1.2/√25 = 0.24. z = (18.4−18)/0.24 = 1.667. p = P(Z>1.667) ≈ 0.0478.
6. p ≈ 0.0478 < 0.05 → reject H₀ (but marginal). Conclusion: Roast time increased slightly.
29. (Defect 2%, n=300, x=11)
1. H₀: p = 0.02
2. Ha: p > 0.02
3. α=0.05, z-test proportion.
4. Reject if p < 0.05.
5. p̂ =11/300=0.036667. SE=sqrt(0.02*0.98/300)=0.007221. z=(0.036667−0.02)/0.007221 ≈
2.33. p=P(Z>2.33)≈0.0099.
6. reject H₀. Conclusion: Defect rate exceeds 2%.
30. (Productivity, n=20, x̄=79, s=6, μ₀=75, α=0.05, right)
1. H₀: μ=75
2. Ha: μ >75
3. t-test df=19.
4. Reject if p<0.05.
5. SE=6/√20≈1.3416. t=(79−75)/1.3416≈2.981. p≈0.004.
6. reject H₀. Conclusion: Productivity increased.
31. (Battery life σ known=2, n=36, x̄=23.2, μ₀=24)
1. H₀: μ = 24
2. Ha: μ < 24
3. α=0.05, z-test (σ known).
4. Reject if p<0.05.
5. SE = 2/√36 = 0.3333. z=(23.2−24)/0.3333 = −2.4. p = P(Z<−2.4) ≈ 0.0082.
6. reject H₀. Conclusion: Battery life less than 24 hrs.
32. (On-time, n=200, x=170)
1. H₀: p = 0.92
2. Ha: p ≠ 0.92
3. α=0.05, z-test proportion.
4. Reject if p<0.05.
5. p̂ =170/200=0.85. SE = sqrt(0.92*0.08/200)=0.0192. z=(0.85−0.92)/0.0192≈−3.646. two-
tailed p≈0.00027.
6. reject H₀. Conclusion: On-time rate differs (lower).
33. (Sodium, n=28, x̄ =2470, s=300, μ₀=2300)
1. H₀: μ=2300
2. Ha: μ > 2300
3. α=0.05, t-test df=27.
4. Reject if p<0.05.
5. SE=300/√28≈56.66. t=(2470−2300)/56.66≈3.0. p≈0.003.
6. reject H₀. Conclusion: Intake increased.
34. (Package weight σ known=3, n=40, x̄=248.7, μ₀=250)
1. H₀: μ=250
2. Ha: μ < 250
3. α=0.05, z-test.
4. Reject if p<0.05.
5. SE=3/√40≈0.4743. z=(248.7−250)/0.4743≈−2.72. p≈0.0033.
6. reject H₀. Conclusion: Underfilling significant.
35. (Employment rate n=120, x=68)
1. H₀: p = 0.60
2. Ha: p ≠ 0.60
3. α=0.05, z-test proportion.
4. Reject if p<0.05.
5. p̂ =68/120≈0.5667. SE=sqrt(0.6*0.4/120)=0.04472. z=(0.5667−0.6)/0.04472≈−0.747.
p≈0.455.
6. fail to reject H₀. Conclusion: No evidence rate differs.
36. (Yield, n=15, x̄ =1.95, s=0.25, μ₀=1.8)
1. H₀: μ=1.8
2. Ha: μ > 1.8
3. α=0.05, t-test df=14.
4. Reject if p<0.05.
5. SE=0.25/√15≈0.06455. t=(1.95−1.8)/0.06455≈2.324. p≈0.018 (one-tailed).
6. reject H₀. Conclusion: Fertilizer seems to increase yield.
37. (Fuel economy, n=22, x̄ =31.6, s=2.8, μ₀=30)
1. H₀: μ=30
2. Ha: μ > 30
3. α=0.05, t-test df=21.
4. Reject if p<0.05.
5. SE=2.8/√22≈0.597. t=1.6/0.597≈2.68. p≈0.007.
6. reject H₀. Conclusion: Economy increased.
38. (Malfunction ≤1%, n=500, x=7)
1. H₀: p = 0.01
2. Ha: p > 0.01
3. α=0.05, z-test proportion.
4. Reject if p<0.05.
5. p̂ =7/500=0.014. SE=sqrt(0.01*0.99/500)=0.00445. z=(0.014−0.01)/0.00445≈0.899.
p≈0.184.
6. fail to reject H₀. Conclusion: Not significantly higher than 1%.
39. (Wait time, n=30, x̄ =9.1, s=2, μ₀=8)
1. H₀: μ=8
2. Ha: μ > 8
3. α=0.05, t-test df=29.
4. Reject if p<0.05.
5. SE=2/√30≈0.3651. t=1.1/0.3651≈3.013. p≈0.0026.
6. reject H₀. Conclusion: Waiting time increased.
40. (Incidents, σ known 1.2, n=36, x̄=3.5, μ₀=3, α=0.05)
1. H₀: μ=3
2. Ha: μ > 3
3. α=0.05, z-test (σ known).
4. Reject if p<0.05.
5. SE=1.2/√36=0.2. z=(3.5−3)/0.2=2.5. p≈0.0062.
6. reject H₀. Conclusion: Incidents increased.
HARD answers (41–60) — tricky (α=0.01 unless specified)
41. (Regulatory claim ≤5%, n=80, x=7, α=0.01)
H₀: p = 0.05 (claim at most 5% — equivalently test p = 0.05)
Ha: p > 0.05 (since audit seeks to see if violations exceed 5%)
z test: p̂ =7/80=0.0875. SE=sqrt(0.05*0.95/80)=0.02437.
z=(0.0875−0.05)/0.02437≈1.558. p = P(Z>1.558) ≈ 0.0595. α=0.01 → p>α → fail to reject
H₀. Conclusion: At 1% level, insufficient evidence that violation rate >5% (but note
p≈0.06 suggests borderline at 5%).
42. (BP increase, n=12, x̄ =128, s=9, μ₀=120, α=0.01, right)
H₀: μ=120; Ha: μ>120. df=11. SE=9/√12≈2.598. t=(128−120)/2.598≈3.078. p≈0.0056.
α=0.01 → p<α → reject H₀. Conclusion: Evidence drug raises BP.
43. (Stadium occupancy, n=150, x=95, p0=0.65, α=0.01)
p̂ = 95/150 ≈ 0.6333. H₀: p=0.65. Ha: p ≠ 0.65. SE = sqrt(0.65*0.35/150)=0.0390.
z=(0.6333−0.65)/0.039≈−0.427. two-tailed p≈0.669. fail to reject H₀.
44. (Precision tool, n=10, x̄ =9.92, s=0.12, μ₀=10, α=0.01)
H₀: μ=10. Ha: μ ≠10 (likely two-sided because “shift” ambiguous). df=9.
SE=0.12/√10≈0.03795. t=(9.92−10)/0.03795≈−2.108. two-tailed p≈0.064. α=0.01 → fail
to reject H₀.
45. (Exam pass, n=500, x=365, p0=0.78, α=0.01)
p̂ =365/500=0.73. H₀: p=0.78; Ha: p ≠ 0.78. SE = sqrt(0.78*0.22/500)=0.0185.
z=(0.73−0.78)/0.0185≈−2.703. two-tailed p≈0.0068. p<0.01 → reject H₀. Conclusion:
Pass rate changed (lower).
46. (Specificity 95%, n=200, false positives=8)
p̂ false pos = 8/200 = 0.04 → specificity observed = 0.96. H₀: specificity = 0.95 →
equivalently p(false) = 0.05. Use proportion test on false rate: H₀: p = 0.05; Ha: p ≠ 0.05.
SE = sqrt(0.05*0.95/200)=0.0154. z=(0.04−0.05)/0.0154≈−0.649. two-tailed p≈0.516. fail
to reject H₀.
47. (Reaction time decreased, n=9, x̄=46.5, s=4.2, μ₀=50, α=0.01 left)
H₀: μ = 50; Ha: μ < 50. df=8. SE = 4.2/√9 = 1.4. t = (46.5−50)/1.4 = −2.5. p = P(T<−2.5,
df=8) ≈ 0.018. α = 0.01 → p>α → fail to reject H₀ at 1% (but significant at 5%).
48. (Energy inefficient, n=120, x=18, p0=0.10, α=0.01)
p̂ =0.15. H₀: p=0.10; Ha: p > 0.10. SE = sqrt(0.1*0.9/120)=0.0274.
z=(0.15−0.10)/0.0274≈1.825. p≈0.034. α=0.01 → fail to reject H₀ (significant at 5%, not
at 1%).
49. (Adverse event, n=300, x=12)
p̂ =0.04. H₀: p=0.02; Ha: p > 0.02. SE = sqrt(0.02*0.98/300)=0.00721.
z=(0.04−0.02)/0.00721≈2.774. p≈0.0027. α=0.01 → reject H₀. Conclusion: adverse rate
increased.
50. (Server response, n=12, x̄ =215, s=15, μ₀=200, α=0.01)
H₀: μ=200; Ha: μ > 200. df=11. SE=15/√12≈4.330. t=(215−200)/4.33≈3.464. p≈0.003
(one-tailed). α=0.01 → reject H₀. Conclusion: response time increased.
51. (Educational improvement, n=10, x̄ =8.2, s=3.5, μ₀=6, α=0.01)
H₀: μ=6; Ha: μ > 6. df=9. SE=3.5/√10≈1.107. t=(8.2−6)/1.107≈1.988. p≈0.038 (one-tailed).
α=0.01 → fail to reject H₀ at 1% (reject at 5%).
52. (Defect standard 4%, n=75, x=5)
p̂ =0.0667. H₀: p=0.04; Ha: p > 0.04. SE = sqrt(0.04*0.96/75)=0.0226.
z=(0.0667−0.04)/0.0226≈1.197. p≈0.115. α=0.01 → fail to reject H₀.
53. (Length of stay, n=11, x̄ =4.2, s=1.1, μ₀=5, α=0.01 left)
H₀: μ=5; Ha: μ < 5. df=10. SE=1.1/√11≈0.3317. t=(4.2−5)/0.3317≈−2.409. p≈0.018.
α=0.01 → fail to reject H₀ (but significant at 5%).
54. (Crash rate, n=1000, x=10, claim 0.005)
Claim crash rate 0.005. H₀: p=0.005; Ha: p > 0.005. p̂=0.01. SE =
sqrt(0.005*0.995/1000)=0.002233. z=(0.01−0.005)/0.002233≈2.239. p≈0.0126. α=0.01
→ fail to reject H₀ (very close).
55. (Clock drift σ known 0.05, n=12, x̄=0.25, μ₀=0.2, α=0.01 right)
H₀: μ=0.2; Ha: μ>0.2. z test: SE = 0.05/√12 ≈ 0.01443. z = (0.25−0.2)/0.01443 ≈ 3.464.
p≈0.00026. reject H₀. Conclusion: drift increased.
56. (Adherence, n=30, x=25, p̂ =0.833)
H₀: p=0.90; Ha: p < 0.90. SE = sqrt(0.9*0.1/30)=0.0548. z=(0.833−0.9)/0.0548≈−1.220.
p≈0.111. α=0.01 → fail to reject H₀.
57. (Yield, n=14, x̄ =3.2, s=0.4, μ₀=3.5, α=0.01 left)
H₀: μ=3.5; Ha: μ<3.5. df=13. SE = 0.4/√14≈0.1069. t=(3.2−3.5)/0.1069≈−2.806. p≈0.0075.
α=0.01 → reject H₀. Conclusion: yield decreased.
58. (Processing time, n=16, x̄=47.8, s=3.7, μ₀=45, α=0.01 right)
H₀: μ=45; Ha: μ>45. df=15. SE=3.7/√16=0.925. t=2.8/0.925≈3.027. p≈0.004. α=0.01 →
reject H₀. Conclusion: processing time longer.
59. (Battery capacity, n=10, x̄=2910, s=120, μ₀=3000, α=0.01 left)
H₀: μ=3000; Ha: μ<3000. df=9. SE=120/√10≈37.95. t=(2910−3000)/37.95≈−2.364.
p≈0.020. α=0.01 → fail to reject H₀ (significant at 5%, not at 1%).
60. (Cycling rate, n=120, x=34, p0=0.30, α=0.01)
p̂ =34/120≈0.2833. H₀: p=0.30; Ha: p > 0.30 (user asked greater) — but here sample
suggests less, so alternative should be chosen by research question. If the advocacy
group wants to test greater: Ha: p>0.30.
z=(0.2833−0.30)/sqrt(0.3*0.7/120)=−0.0167/0.0428≈−0.391. p≈0.652. fail to reject H₀. If
two-sided, also fail to reject.