1. What is the primary purpose of calculating sample size before starting a research study?
A. To reduce publication time
B. To determine the number of subjects required for a valid study
C. To avoid statistical analysis
D. To increase study cost
Answer: B
2. Which of the following is one of the objectives of this lecture?
A. Developing questionnaires
B. Understanding the relationship between sample size and power
C. Writing research papers
D. Conducting systematic reviews
Answer: B
3. Which concept is closely related to sample size in biomedical research?
A. Blinding
B. Randomization
C. Power
D. Confounding
Answer: C
4. According to the lecture, is there a simple answer to the question “How many subjects are required?”
A. Yes, always 100
B. Yes, always 500
C. No, it depends on several study factors
D. Always determined by the supervisor
Answer: C
5. The first step in estimating sample size is to:
A. Select statistical software
B. Identify the major study variable
C. Recruit participants
D. Calculate confidence intervals
Answer: B
6. The major study variable refers to:
A. The least important variable
B. The variable of greatest interest in the research
C. Any demographic variable
D. The study title
Answer: B
7. If a study aims to estimate the prevalence of scrub typhus, the major study variable is:
A. Age
B. Gender
C. Presence or absence of scrub typhus
D. Education level
Answer: C
8. After identifying the major study variable, the next step is to determine the:
A. Study budget
B. Type of estimate
C. Journal for publication
D. Sample collection method
Answer: B
9. Which of the following is an example of a type of estimate?
A. Mean
B. Proportion
C. Ratio
D. All of the above
Answer: D
10. Expected frequency of the factor of interest mainly affects the:
A. Font size of the report
B. Required sample size
C. Research title
D. Ethics approval
Answer: B
11. A very rare disease generally requires:
A. A very small sample
B. No sample
C. A larger sample
D. One participant
Answer: C
12. A very common disease usually requires:
A. A larger sample than rare diseases
B. A smaller sample
C. No sample calculation
D. Only pilot data
Answer: B
13. Precision refers to:
A. Study duration
B. Closeness of the estimate to the true value
C. Sample collection
D. Disease frequency
Answer: B
14. If higher precision is required, the sample size should generally be:
A. Smaller
B. Larger
C. Unchanged
D. Zero
Answer: B
15. Which precision requires a larger sample size?
A. ±20%
B. ±15%
C. ±10%
D. ±5%
Answer: D
16. Acceptable risk that an estimate falls outside the true population value is related to:
A. Confidence level
B. Height
C. Income
D. Weight
Answer: A
17. Sample size formulas generally assume sampling from:
A. A very small population
B. A very large population
C. One household
D. Only hospital patients
Answer: B
18. If the population is small, sample size should be adjusted using:
A. Population correction
B. Randomization
C. Matching
D. Stratification
Answer: A
19. Which adjustment is needed when cluster sampling is used?
A. Precision adjustment
B. Design effect adjustment
C. Income adjustment
D. Matching adjustment
Answer: B
20. Design effect mainly accounts for:
A. Measurement error
B. Correlation among subjects within clusters
C. Publication bias
D. Recall bias
Answer: B
21. Expected response rate is considered because:
A. Some participants may not respond
B. Journals require it
C. It increases publication chances
D. It changes study objectives
Answer: A
22. If 10% non-response is expected, investigators should:
A. Reduce the sample size
B. Increase the calculated sample size
C. Ignore non-response
D. Stop the study
Answer: B
23. Which of the following is NOT a step in estimating sample size?
A. Identify the major study variable
B. Decide desired precision
C. Decide acceptable risk
D. Select the journal before data collection
Answer: D
24. Which statement about sample size estimation is TRUE?
A. One formula applies to every research study.
B. Sample size depends on study objectives and statistical parameters.
C. Sample size is always 100.
D. Sample size does not influence study quality.
Answer: B
25. Which statement BEST summarizes the initial steps of sample size estimation?
A. Estimating sample size begins by identifying the major study variable, determining the type of estimate (mean, proportion, ratio), estimating the expected frequency, deciding the desired precision and acceptable confidence level, and making adjustments for population size, design effect, and expected non-response to ensure adequate statistical validity.
B. Sample size depends only on the investigator’s preference.
C. Every study requires exactly the same number of participants.
D. Sample size is calculated only after data collection.
Answer: A
26. Alpha (α) represents:
A. Probability of making a Type II error
B. Probability of rejecting the null hypothesis when it is true
C. Study power
D. Sample size
Answer: B
27. Type I error occurs when:
A. A false null hypothesis is accepted
B. A true null hypothesis is rejected
C. The sample size is too large
D. Randomization fails
Answer: B
28. Alpha (α) is also known as the:
A. Confidence coefficient
B. Significance level
C. Power
D. Precision
Answer: B
29. Confidence level is expressed as:
A. α
B. β
C. 1 − α
D. 1 − β
Answer: C
30. A 95% confidence level corresponds approximately to an alpha value of:
A. 0.50
B. 0.10
C. 0.05
D. 0.01
Answer: C
31. If α = 0.05, the confidence level is:
A. 80%
B. 90%
C. 95%
D. 99%
Answer: C
32. Beta (β) represents:
A. Probability of rejecting a true null hypothesis
B. Probability of failing to reject a false null hypothesis
C. Confidence level
D. Precision
Answer: B
33. Beta (β) is also known as:
A. Type I error
B. Type II error
C. Sampling error
D. Random error
Answer: B
34. Power of a study is defined as:
A. α
B. β
C. 1 − β
D. 1 − α
Answer: C
35. Study power represents the probability of:
A. Accepting a true null hypothesis
B. Correctly rejecting a false null hypothesis
C. Rejecting every hypothesis
D. Calculating sample size
Answer: B
36. Which pair is correctly matched?
A. α → Type II error
B. β → Confidence level
C. 1 − β → Power
D. 1 − α → Type II error
Answer: C
37. Increasing study power generally requires:
A. Smaller sample size
B. Larger sample size
C. Lower confidence level
D. Fewer participants
Answer: B
38. Precision is best defined as:
A. Difference between two means
B. Closeness of an estimate to the true population value
C. Disease prevalence
D. Type II error
Answer: B
39. Precision may be expressed as:
A. Absolute terms only
B. Relative terms only
C. Absolute or relative terms
D. Percentages only
Answer: C
40. When estimating a population mean, the required sample size depends on:
A. Population standard deviation
B. Desired precision
C. Confidence level
D. All of the above
Answer: D
41. Which symbol represents the population standard deviation?
A. μ
B. σ
C. β
D. λ
Answer: B
42. Which value is commonly used as the Z-score for a 95% confidence interval?
A. 1.28
B. 1.64
C. 1.96
D. 2.58
Answer: C
43. The standard error of the sample mean is calculated as:
A. σ × √n
B. σ / √n
C. √σ
D. n / σ
Answer: B
44. Which formula is used to estimate sample size for estimating a population mean?
A. n = Z²σ² / d²
B. n = Z²pq / d²
C. n = p/q
D. n = σ/d
Answer: A
45. In the sample size formula for estimating a mean, the symbol ‘d’ represents:
A. Disease prevalence
B. Desired precision (margin of error)
C. Design effect
D. Difference between groups
Answer: B
46. Which factor directly increases the required sample size for estimating a population mean?
A. Smaller standard deviation
B. Larger desired margin of error
C. Greater population variability (σ)
D. Lower confidence level
Answer: C
47. In the lecture example of estimating average daily protein intake, the estimated population standard deviation was:
A. 10 g
B. 15 g
C. 20 g
D. 25 g
Answer: C
48. In the same example, the desired precision (d) was:
A. 2 units
B. 5 units
C. 10 units
D. 20 units
Answer: B
49. The calculated sample size for estimating the average daily protein intake of teenage girls was approximately:
A. 30
B. 45
C. 62
D. 120
Answer: C
50. Which statement BEST summarizes the concepts of alpha, beta, power, precision, and sample size estimation for a population mean?
A. Alpha (α) is the probability of making a Type I error, while beta (β) is the probability of making a Type II error. Confidence level equals 1 − α, and study power equals 1 − β. Sample size for estimating a population mean depends on the desired confidence level, precision, and population standard deviation, using the formula n = Z²σ²/d². Increasing confidence or precision generally requires a larger sample size.
B. Alpha and beta represent the same statistical concept and are interchangeable.
C. Power is equal to alpha, and confidence level is equal to beta.
D. Sample size for estimating a mean depends only on the investigator’s preference.
Answer: A
51. Which formula is commonly used to estimate the sample size for a population proportion?
A. n = Z²σ²/d²
B. n = Z²pq/d²
C. n = p/q
D. n = Z/d
Answer: B
52. In the formula n = Z²pq/d², the symbol p represents:
A. Precision
B. Population proportion (expected prevalence)
C. Population variance
D. Power
Answer: B
53. In the same formula, q is equal to:
A. p
B. 1 + p
C. 1 − p
D. p²
Answer: C
54. When the expected prevalence (p) is unknown, the lecture recommends using:
A. 0.1
B. 0.25
C. 0.5
D. 0.9
Answer: C
55. Why is p = 0.5 commonly used when no estimate is available?
A. It gives the smallest sample size.
B. It gives the maximum sample size.
C. It eliminates sampling error.
D. It increases study power to 100%.
Answer: B
56. The expected value of p can be obtained from:
A. Previous studies
B. Pilot studies
C. Published literature
D. All of the above
Answer: D
57. In the lecture example of estimating immunization coverage, the expected proportion (p) was:
A. 0.20
B. 0.50
C. 0.80
D. 0.95
Answer: C
58. In the same example, the absolute precision (d) was:
A. 0.02
B. 0.04
C. 0.08
D. 0.10
Answer: B
59. The calculated sample size for estimating immunization coverage was approximately:
A. 100
B. 247
C. 384
D. 500
Answer: C
60. Why was the calculated sample size increased after estimation in the lecture example?
A. To improve publication quality
B. To adjust for design effect
C. To reduce alpha
D. To eliminate bias completely
Answer: B
61. The Design Effect (DEFF) is mainly used when:
A. Simple random sampling is used
B. Cluster sampling is used
C. Case reports are prepared
D. Laboratory studies are performed
Answer: B
62. Design effect accounts for:
A. Recall bias
B. Correlation among individuals within the same cluster
C. Publication bias
D. Measurement error
Answer: B
63. If the design effect is greater than 1, the required sample size will generally:
A. Decrease
B. Remain unchanged
C. Increase
D. Become zero
Answer: C
64. Which adjustment should be made when a response rate lower than 100% is expected?
A. Reduce the sample size
B. Increase the sample size
C. Ignore non-response
D. Change the study design
Answer: B
65. Why is adjustment for non-response necessary?
A. To improve journal acceptance
B. To ensure the final completed sample remains adequate
C. To reduce confidence level
D. To eliminate sampling error
Answer: B
66. Which factor has the greatest influence on sample size when estimating a proportion?
A. Investigator’s experience
B. Expected prevalence, desired precision, and confidence level
C. Gender distribution
D. Study location only
Answer: B
67. A researcher wants a narrower confidence interval. What should be done?
A. Reduce the sample size
B. Increase the desired precision
C. Increase the sample size
D. Increase beta
Answer: C
68. Which statement regarding sample size estimation for proportions is TRUE?
A. Smaller precision requires fewer participants.
B. Greater precision requires more participants.
C. Confidence level has no effect on sample size.
D. Population proportion is unnecessary.
Answer: B
69. Which parameter is represented by Z in the sample size formula?
A. Standard deviation
B. Confidence level from the normal distribution
C. Disease prevalence
D. Study power
Answer: B
70. If a researcher changes the confidence level from 95% to 99%, the required sample size will generally:
A. Decrease
B. Remain the same
C. Increase
D. Become half
Answer: C
71. Which of the following BEST explains the relationship between precision and sample size?
A. Higher precision requires a larger sample.
B. Precision is unrelated to sample size.
C. Lower precision always increases sample size.
D. Precision depends only on prevalence.
Answer: A
72. A pilot study is mainly useful because it helps estimate:
A. Journal impact factor
B. Population parameters needed for sample size calculation
C. Ethical approval
D. Randomization sequence
Answer: B
73. If previous literature provides an estimate of prevalence, that estimate can be used to determine:
A. Sampling frame
B. Expected proportion (p)
C. Type I error
D. Odds ratio
Answer: B
74. Which statement regarding sample size estimation is CORRECT?
A. Different study objectives require different sample size formulas.
B. One formula applies to every study design.
C. Sample size depends only on disease prevalence.
D. Sample size is independent of confidence level.
Answer: A
75. Which statement BEST summarizes sample size estimation for proportions?
A. When estimating a population proportion, the sample size depends on the expected prevalence (p), its complement (q = 1 − p), the desired precision (d), and the confidence level (Z). If prevalence is unknown, p = 0.5 is commonly used because it produces the largest sample size. The calculated sample should then be adjusted for design effect and expected non-response whenever applicable.
B. Sample size for proportions depends only on the investigator’s preference.
C. The expected prevalence is never required for sample size estimation.
D. Design effect reduces the required sample size in cluster sampling.
Answer: A
76. Which information is essential before calculating the sample size for a cohort study?
A. Participant names
B. Baseline disease proportion and expected relative risk
C. Study budget only
D. Journal impact factor
Answer: B
77. For analytical studies such as cohort or case-control studies, sample size calculation requires knowledge of:
A. Alpha (α)
B. Beta (β)
C. Expected effect size
D. All of the above
Answer: D
78. In a cohort study, the expected effect size is commonly expressed as:
A. Correlation coefficient
B. Relative Risk (RR)
C. Standard deviation
D. Mean difference only
Answer: B
79. In a case-control study, the expected effect size is usually expressed as:
A. Relative Risk
B. Odds Ratio (OR)
C. Correlation coefficient
D. Hazard Ratio only
Answer: B
80. Previous studies reported that 15% of non-oral contraceptive users develop myocardial infarction, while 25% of users develop it. This information is mainly used to estimate:
A. Precision
B. Sample size for a cohort study
C. Study duration
D. Response rate
Answer: B
81. In the lecture example of the cohort study, the desired study power was:
A. 50%
B. 60%
C. 80%
D. 99%
Answer: C
82. The conventional beta (β) value used in the cohort study example was:
A. 0.01
B. 0.05
C. 0.20
D. 0.50
Answer: C
83. In the cohort study example, the required sample size for each group was approximately:
A. 100
B. 247
C. 384
D. 500
Answer: B
84. Equal sample size in exposed and non-exposed groups mainly improves:
A. Simplicity and statistical efficiency
B. Publication quality
C. Ethics approval
D. Response rate
Answer: A
85. In the lecture’s case-control study example, the expected Odds Ratio (OR) was:
A. 1.2
B. 1.5
C. 1.8
D. 3.0
Answer: C
86. The calculated sample size for the case-control study was approximately:
A. 150 cases and controls
B. 247 cases and controls
C. 292 cases and 292 controls
D. 384 cases and controls
Answer: C
87. Detecting a very small Odds Ratio (e.g., OR = 1.2) generally requires:
A. A smaller sample
B. No sample calculation
C. A much larger sample
D. Only pilot data
Answer: C
88. Detecting a large Odds Ratio (e.g., OR = 3) generally requires:
A. A much larger sample
B. A smaller sample
C. No statistical analysis
D. Cluster sampling only
Answer: B
89. Which statement best explains the relationship between effect size and sample size?
A. Smaller effect sizes require larger samples.
B. Larger effect sizes require larger samples.
C. Effect size has no influence on sample size.
D. Sample size depends only on alpha.
Answer: A
90. A confounder is best described as:
A. A random error
B. A third variable affecting the observed association
C. A sampling method
D. A significance level
Answer: B
91. According to the lecture, sample size should generally be increased by approximately:
A. 2% per confounder
B. 5% per confounder
C. 10% per confounder
D. 20% per confounder
Answer: C
92. Which free software supported by the CDC is recommended for sample size calculation?
A. SPSS
B. OpenEpi
C. GraphPad Prism
D. Stata
Answer: B
93. Which free software developed by Vanderbilt University is recommended in the lecture?
A. Epi Info
B. PS (Power and Sample Size)
C. RStudio
D. RevMan
Answer: B
94. When reporting sample size in a research paper, investigators should mention:
A. The software used
B. The assumptions entered into the software
C. The calculated sample size
D. All of the above
Answer: D
95. Sample size calculation should be reported in which section of a research paper?
A. Introduction
B. Results
C. Methods
D. References
Answer: C
96. A researcher wants to detect a very small treatment effect with 95% confidence and 90% power. Which of the following is most likely?
A. A very small sample will be sufficient.
B. A larger sample size will be required.
C. Sample size will decrease.
D. Precision becomes unnecessary.
Answer: B
97. A community survey uses cluster sampling and expects a 15% non-response rate. Which adjustments should be applied after the initial sample size calculation?
A. Design effect only
B. Non-response adjustment only
C. Both design effect and non-response adjustment
D. No adjustment is necessary
Answer: C
98. A researcher has no previous estimate of disease prevalence but wants to estimate prevalence with maximum statistical safety. Which value of p should be used?
A. 0.10
B. 0.25
C. 0.50
D. 0.90
Answer: C
99. A study aims to detect an Odds Ratio of only 1.2 instead of 3.0. Compared with OR = 3.0, the required sample size will be:
A. Much smaller
B. Nearly the same
C. Much larger
D. Zero
Answer: C
100. Which statement BEST summarizes the lecture on Calculating Sample Size and Power?
A. There is no universal sample size for research. The required sample depends on the study objective, major study variable, type of estimate, expected prevalence or variability, desired precision, confidence level (α), power (1−β), study design, anticipated effect size, design effect, expected non-response, and possible confounders. Different formulas are used for estimating means, proportions, cohort studies, and case-control studies. Modern software such as OpenEpi and PS simplifies these calculations, and all assumptions used should be clearly reported in the Methods section of the research report.
B. Every biomedical study requires exactly 384 participants regardless of study design.
C. Sample size depends only on confidence level because all other factors have little effect.
D. Once a sample size is calculated, no adjustments for non-response, design effect, or confounders are ever necessary.
Answer: A
