1. The primary goal of an epidemiological study is to:
A. Increase disease prevalence
B. Obtain accurate estimates of disease frequency or exposure–outcome relationships
C. Reduce healthcare expenditure
D. Develop new drugs
Answer: B
2. Internal validity refers to:
A. Generalizing study findings to other populations
B. Accuracy of the study results within the study population
C. Cost-effectiveness of the study
D. Sample size adequacy
Answer: B
3. External validity refers to:
A. Accuracy of laboratory instruments
B. Precision of measurements
C. Generalizability of study findings to the target population
D. Random error
Answer: C
4. Which of the following is essential before study findings can be generalized?
A. External validity only
B. Internal validity
C. Large sample size
D. Randomization
Answer: B
5. The term accuracy in epidemiological research consists of:
A. Validity and reliability
B. Precision and validity
C. Precision and power
D. Sensitivity and specificity
Answer: B
6. Precision refers to:
A. Closeness of results to the true value
B. Consistency or reproducibility of repeated measurements
C. Absence of confounding
D. Generalizability
Answer: B
7. Validity refers to:
A. Repeatability of observations
B. Closeness of measurements to the true value
C. Sample representativeness only
D. Statistical significance
Answer: B
8. A study repeatedly produces nearly identical results, but all results are far from the true value. The study is:
A. Precise but not valid
B. Valid but not precise
C. Both valid and precise
D. Neither valid nor precise
Answer: A
9. Which combination represents the ideal epidemiological study?
A. Valid but imprecise
B. Precise but invalid
C. Precise and valid
D. Neither precise nor valid
Answer: C
10. Which of the following is the worst-case scenario?
A. High precision and high validity
B. High precision and low validity
C. Low precision and high validity
D. Low precision and low validity
Answer: D
11. Random error occurs mainly because of:
A. Bias
B. Chance variation
C. Confounding
D. Selection procedures
Answer: B
12. Which of the following is an example of random error?
A. Sampling variability
B. Selection bias
C. Recall bias
D. Observer bias
Answer: A
13. Systematic error is also known as:
A. Random error
B. Precision error
C. Bias
D. Sampling variation
Answer: C
14. Which type of error poses the greatest threat to study validity?
A. Random error
B. Chance variation
C. Systematic error (bias)
D. Sampling fluctuation
Answer: C
15. Bias can best be defined as:
A. Random variation in observations
B. A process producing results that systematically deviate from the truth
C. Small sample size
D. Poor statistical analysis only
Answer: B
16. Which of the following is NOT considered a major threat to validity?
A. Selection bias
B. Information bias
C. Confounding
D. Precision
Answer: D
17. The three major threats to validity in epidemiological studies are:
A. Chance, prevalence, incidence
B. Selection bias, information bias, confounding
C. Sensitivity, specificity, predictive value
D. Reliability, reproducibility, repeatability
Answer: B
18. Selection bias occurs primarily because:
A. Exposure is measured inaccurately.
B. Participants selected do not appropriately represent the target population.
C. Laboratory instruments are poorly calibrated.
D. Confounders are not adjusted.
Answer: B
19. Which question best helps identify selection bias?
A. Were laboratory tests standardized?
B. Do study participants accurately represent the target population?
C. Were measurements repeated?
D. Was regression analysis performed?
Answer: B
20. Selection bias mainly affects:
A. Participant selection
B. Statistical software
C. Laboratory quality control
D. Randomization sequence
Answer: A
21. Information bias primarily results from:
A. Incorrect participant selection
B. Errors in measuring exposure, outcome, or other variables
C. Small sample size
D. Loss to follow-up
Answer: B
22. Confounding occurs when:
A. Sampling variability increases.
B. A third factor distorts the true association between exposure and outcome.
C. Participants refuse consent.
D. Randomization fails.
Answer: B
23. Which of the following best describes a confounder?
A. A laboratory error
B. A third variable associated with both exposure and outcome
C. A measurement instrument
D. A source of random error
Answer: B
24. Which statement regarding validity is CORRECT?
A. A study with poor internal validity can still have excellent external validity.
B. Internal validity is necessary before external validity can be achieved.
C. External validity is more important than internal validity.
D. Precision guarantees validity.
Answer: B
25. Which statement BEST summarizes validity in epidemiological studies?
A. An epidemiological study should first ensure accurate, unbiased estimates within the study population (internal validity) before its findings can be generalized to the target population (external validity), while minimizing random error, bias, and confounding.
B. External validity is independent of internal validity.
C. Precision alone determines study quality.
D. Large sample size eliminates all threats to validity.
Answer: A
26. Which of the following is the most common cause of selection bias?
A. Incorrect laboratory calibration
B. Improper selection of study participants
C. Data entry error
D. Statistical software error
Answer: B
27. Selection bias occurs when:
A. Exposure is measured incorrectly
B. Study participants do not accurately represent the target population
C. Disease is misclassified
D. Laboratory results vary randomly
Answer: B
28. In a case-control study, selection bias may occur if:
A. Cases are selected from the community and controls from the same community
B. Cases are selected from hospitals while controls represent a different source population
C. Random sampling is used
D. Blinding is performed
Answer: B
29. Which situation is an example of surveillance bias?
A. Exposed individuals are monitored more intensively than unexposed individuals.
B. Equal follow-up of all participants.
C. Random allocation of treatment.
D. Blinded outcome assessment.
Answer: A
30. Screening bias occurs when:
A. Exposed individuals are more likely to undergo disease detection than unexposed individuals.
B. Participants refuse consent.
C. Sample size is too small.
D. Measurements are repeated.
Answer: A
31. Admission rate bias is MOST likely in studies where participants are selected from:
A. Community surveys
B. Hospitals and healthcare facilities
C. National census data
D. Birth registries
Answer: B
32. Which type of selection bias occurs when only surviving cases are included in the study?
A. Recall bias
B. Survival (survivor) bias
C. Interviewer bias
D. Observer bias
Answer: B
33. In cohort studies, selection bias commonly results from:
A. Misclassification of exposure
B. Loss to follow-up
C. Recall bias
D. Randomization
Answer: B
34. Loss to follow-up introduces bias mainly when:
A. Dropouts occur equally in all exposure groups.
B. Dropouts differ systematically according to exposure or disease risk.
C. Sample size increases.
D. Blinding is maintained.
Answer: B
35. Which strategy helps reduce survival bias in case-control studies?
A. Using prevalent cases
B. Using incident cases
C. Increasing sample size
D. Matching only
Answer: B
36. Which design is generally LESS prone to selection bias in case-control studies?
A. Hospital-based design
B. Population-based design
C. Convenience sampling
D. Volunteer sampling
Answer: B
37. Applying identical eligibility criteria to both cases and controls primarily reduces:
A. Recall bias
B. Selection bias
C. Confounding
D. Random error
Answer: B
38. Which practice helps minimize selection bias during participant identification?
A. Different diagnostic criteria for cases and controls
B. Identical diagnostic procedures for all participants
C. Different screening intensity
D. Self-reported diagnosis only
Answer: B
39. During data collection, which strategy BEST minimizes selection bias?
A. Maximizing non-response
B. Minimizing non-response and loss to follow-up
C. Increasing recall period
D. Using convenience sampling
Answer: B
40. Why should investigators record baseline characteristics of participants lost to follow-up?
A. To calculate prevalence
B. To compare them with retained participants and assess possible selection bias
C. To increase precision
D. To estimate incidence
Answer: B
41. Blinding during participant selection primarily helps prevent:
A. Measurement error
B. Selection bias related to exposure knowledge
C. Confounding
D. Random error
Answer: B
42. At the analysis stage, investigators compare responders with non-responders mainly to:
A. Increase statistical power
B. Detect possible selection bias
C. Reduce recall bias
D. Estimate prevalence
Answer: B
43. Large baseline differences between responders and non-responders usually suggest:
A. Confounding only
B. Selection bias
C. Random error
D. Effect modification
Answer: B
44. Small differences between responders and non-responders:
A. Completely rule out selection bias.
B. Reduce sample size.
C. Do not completely exclude selection bias.
D. Eliminate confounding.
Answer: C
45. Sensitivity analysis is performed to:
A. Diagnose diseases
B. Assess the potential impact and direction of selection bias
C. Increase sample size
D. Improve laboratory precision
Answer: B
46. Information bias primarily results from:
A. Improper participant selection
B. Systematic errors in measuring exposure, outcome, or other variables
C. Random sampling variation
D. Small sample size
Answer: B
47. Which variables are commonly measured in epidemiological studies and therefore susceptible to information bias?
A. Exposure only
B. Outcome only
C. Exposure, outcome, and other relevant variables (confounders/modifiers)
D. Sample size only
Answer: C
48. The main objective while collecting epidemiological data is to ensure that measurements:
A. Increase disease prevalence.
B. Accurately represent the true exposure and outcome.
C. Reduce external validity.
D. Eliminate randomization.
Answer: B
49. Which of the following BEST defines information bias?
A. Bias arising from participant selection.
B. Bias caused by systematic errors in collecting or measuring study information.
C. Bias resulting from confounding variables.
D. Bias caused by sampling variability.
Answer: B
50. Which statement BEST summarizes the prevention of selection and information bias?
A. Selection bias is minimized through appropriate participant selection, equal eligibility criteria, standardized diagnostic procedures, minimizing loss to follow-up, and sensitivity analysis, while information bias is reduced through accurate and standardized measurement of exposures, outcomes, and other variables.
B. Increasing sample size alone prevents both biases.
C. Randomization eliminates all information bias.
D. Blinding alone prevents every type of bias.
Answer: A
51. Recall bias is most commonly associated with:
A. Cross-sectional studies
B. Cohort studies
C. Case-control studies
D. Clinical trials
Answer: C
52. Recall bias occurs because:
A. Investigators forget to record data.
B. Cases remember previous exposures differently from controls.
C. Laboratory instruments are inaccurate.
D. Participants are randomly selected.
Answer: B
53. In a case-control study, information bias may occur when:
A. Exposure history is collected more thoroughly from cases than controls.
B. Randomization is inadequate.
C. Sample size is large.
D. Cases and controls are selected from the same population.
Answer: A
54. Which of the following is an example of information bias?
A. Hospital admission bias
B. Recall bias
C. Survival bias
D. Volunteer bias
Answer: B
55. In a cohort study, information bias may occur if:
A. Exposed participants are followed more carefully than unexposed participants.
B. Participants are randomized.
C. Incident cases are selected.
D. Matching is performed.
Answer: A
56. Better outcome data available only among exposed participants can lead to:
A. Selection bias
B. Information bias
C. Confounding
D. Random error
Answer: B
57. Which of the following may introduce information bias by the investigator?
A. Systematic collection of data supporting the expected hypothesis
B. Equal measurement of all participants
C. Standardized protocol
D. Blinded assessment
Answer: A
58. Participants intentionally providing incorrect information during interviews is known as:
A. Recall bias
B. Observer bias
C. Prevarication
D. Selection bias
Answer: C
59. Which is the FIRST step in preventing information bias?
A. Multivariable analysis
B. Precise operational definitions of variables
C. Increasing sample size
D. Randomization
Answer: B
60. Detailed measurement protocols primarily improve:
A. Internal validity
B. External validity only
C. Disease prevalence
D. Sample size
Answer: A
61. Repeated measurements of blood pressure are recommended mainly to:
A. Increase disease incidence
B. Reduce measurement error
C. Produce selection bias
D. Increase confounding
Answer: B
62. Training and certification of investigators primarily reduce:
A. Information bias
B. Random error only
C. Selection bias only
D. Confounding
Answer: A
63. Data audits in epidemiological studies are mainly performed to:
A. Increase response rate
B. Ensure data are collected, stored, and managed correctly
C. Reduce prevalence
D. Perform matching
Answer: B
64. Data cleaning before analysis is important because it:
A. Eliminates confounding completely.
B. Detects inconsistencies and improves data quality.
C. Reduces sample size.
D. Removes randomization.
Answer: B
65. Re-running statistical analyses before publication helps to:
A. Increase disease frequency.
B. Detect analytical errors that may introduce information bias.
C. Eliminate recall bias.
D. Prevent loss to follow-up.
Answer: B
66. Confounding is best described as:
A. Random sampling variation
B. Confusion of effects caused by a third variable
C. Laboratory measurement error
D. Poor participant selection
Answer: B
67. A confounder must be associated with:
A. Exposure only
B. Outcome only
C. Both exposure and outcome
D. Neither exposure nor outcome
Answer: C
68. Which statement regarding confounding is TRUE?
A. It always weakens an association.
B. It affects only randomized trials.
C. It can distort the true exposure–outcome relationship.
D. It is a type of random error.
Answer: C
69. Confounding may produce:
A. A false association when none exists
B. A weaker association
C. A stronger association
D. All of the above
Answer: D
70. Which of the following is considered the greatest threat to validity in epidemiological studies?
A. Random error
B. Information bias
C. Confounding
D. Precision
Answer: C
71. Confounding can hide:
A. True associations
B. Laboratory errors
C. Sampling errors
D. Measurement protocols
Answer: A
72. Which of the following is the MOST dangerous consequence of confounding?
A. Reduced precision
B. Increased sample size
C. Reversal of the direction of an observed association
D. Increased response rate
Answer: C
73. Restriction is used to control confounding by:
A. Selecting participants from only one level of a confounding variable
B. Increasing follow-up duration
C. Blinding investigators
D. Increasing sample size
Answer: A
74. Matching controls confounding by:
A. Measuring exposure repeatedly
B. Making cases and controls similar with respect to potential confounders
C. Increasing precision
D. Reducing recall bias
Answer: B
75. Which statement BEST summarizes information bias and confounding?
A. Information bias results from systematic errors in measuring exposures, outcomes, or other variables and can be minimized through standardized measurement procedures, whereas confounding occurs when a third variable distorts the true exposure–outcome relationship and can create, hide, strengthen, weaken, or even reverse an observed association.
B. Information bias and confounding are identical.
C. Randomization completely eliminates information bias.
D. Increasing sample size removes confounding.
Answer: A
76. A case-control study reports that smoking is associated with lung cancer. Later analysis shows that occupational asbestos exposure influenced both smoking habits and lung cancer risk. This is an example of:
A. Selection bias
B. Recall bias
C. Confounding
D. Random error
Answer: C
77. Which of the following is the BEST design-stage method to control confounding in a randomized clinical trial?
A. Stratification
B. Multivariable regression
C. Randomization
D. Sensitivity analysis
Answer: C
78. Restriction controls confounding by:
A. Measuring the confounder repeatedly
B. Including participants from only one category of a potential confounder
C. Increasing the follow-up period
D. Blinding investigators
Answer: B
79. Matching is most commonly used in:
A. Cross-sectional studies
B. Ecological studies
C. Case-control studies
D. Case series
Answer: C
80. When matching is used during study design, which type of statistical analysis should generally be performed?
A. Crude analysis only
B. Matched analysis
C. Survival analysis only
D. Time-series analysis
Answer: B
81. Which of the following is an analysis-stage method to control confounding?
A. Restriction
B. Randomization
C. Stratification
D. Blinding
Answer: C
82. Multivariable analysis is mainly performed to:
A. Increase disease prevalence
B. Adjust for multiple confounding variables simultaneously
C. Eliminate random error
D. Improve participant recruitment
Answer: B
83. Which statistical technique is commonly used in multivariable analysis?
A. Logistic regression
B. Linear regression
C. Multiple regression models
D. All of the above
Answer: D
84. Which statement regarding confounding control is TRUE?
A. It is always easier to control confounding during analysis than during study design.
B. Confounding is best prevented during the design stage whenever possible.
C. Confounding cannot be controlled.
D. Only randomization controls confounding.
Answer: B
85. Before concluding that an observed association is causal, researchers should first exclude:
A. Chance
B. Bias
C. Confounding
D. All of the above
Answer: D
86. A researcher observes a significant crude odds ratio between alcohol intake and liver disease. After adjusting for hepatitis infection, the association disappears. This suggests:
A. Random error
B. Confounding
C. Recall bias
D. Observer bias
Answer: B
87. Which statement best describes a crude association?
A. The true causal association after all adjustments
B. An observed association before evaluating chance, bias, and confounding
C. An association obtained only from randomized trials
D. A statistically significant association
Answer: B
88. According to epidemiological principles, the pathway from a crude association to a causal association requires evaluation of:
A. Chance only
B. Bias only
C. Confounding only
D. Chance, bias, and confounding
Answer: D
89. Which of the following should be evaluated FIRST when interpreting epidemiological results?
A. Whether the association is statistically significant
B. Whether the findings may be explained by chance
C. Whether the journal has a high impact factor
D. Whether the sample size is large
Answer: B
90. A study finds that coffee consumption doubles the risk of myocardial infarction. Before accepting this conclusion, investigators should evaluate:
A. Selection bias
B. Information bias
C. Confounding by smoking
D. All of the above
Answer: D
91. In the coffee and heart attack example, smoking is considered:
A. An outcome
B. A confounder
C. A source of random error
D. A selection criterion
Answer: B
92. Smoking acts as a confounder because it is associated with:
A. Coffee consumption only
B. Heart attack only
C. Both coffee consumption and heart attack
D. Neither exposure nor outcome
Answer: C
93. A hospital-based coffee study enrolls patients admitted with gastric ulcers. This primarily increases the risk of:
A. Recall bias
B. Selection bias
C. Measurement error
D. Effect modification
Answer: B
94. Participants inaccurately reporting the number of cups of coffee consumed each day is an example of:
A. Information bias
B. Selection bias
C. Randomization error
D. Matching bias
Answer: A
95. Misdiagnosing chest pain as myocardial infarction in a study represents:
A. Selection bias
B. Information bias
C. Confounding
D. Random error only
Answer: B
96. Which statement regarding observational studies is TRUE?
A. They are generally less susceptible to bias than randomized trials.
B. Biases occur only during data analysis.
C. They are particularly vulnerable to selection bias, information bias, and confounding.
D. Confounding cannot occur.
Answer: C
97. Biases may occur during:
A. Study design only
B. Data collection only
C. Data analysis only
D. Study design, implementation, and analysis
Answer: D
98. Which statement about internal and external validity is CORRECT?
A. External validity can exist without internal validity.
B. Internal validity is a prerequisite for external validity.
C. External validity is determined only by sample size.
D. Internal validity depends only on statistical significance.
Answer: B
99. Even when unavoidable biases remain, investigators should:
A. Ignore them if results are statistically significant.
B. Mention them as study limitations.
C. Remove affected participants after analysis.
D. Repeat the entire study immediately.
Answer: B
100. Which statement BEST summarizes the evaluation of validity in epidemiological studies?
A. Valid interpretation of epidemiological findings requires first ensuring internal validity through minimizing random error, selection bias, information bias, and confounding. Only after these threats have been adequately addressed can the observed association be considered causal and generalized to the target population.
B. Statistical significance alone proves causality.
C. External validity is more important than internal validity.
D. Large sample size eliminates all threats to validity.
Answer: A
