These questions cover the Day 2 topics: Correlation and Regression, Survival Analysis, Screening Tests, and Clinical Trial Design.
Which of the following statements about the correlation coefficient are TRUE?
Correct answers: A, C, D, E are TRUE
Explanation:
A) TRUE – By definition, r can range from -1 (perfect negative correlation) to +1 (perfect positive correlation).
B) FALSE – A correlation of -0.8 indicates a strong negative association. Values close to -1 or +1 indicate strong correlations; values close to 0 indicate weak correlations.
C) TRUE – When r = 0, there is no linear relationship between the variables. However, there may still be a non-linear relationship.
D) TRUE – The correlation coefficient has no units; it is a pure number describing the strength and direction of linear association.
E) TRUE – The correlation between x and y is the same as between y and x. The order of variables does not affect the coefficient.
Which of the following statements about Pearson’s and Spearman’s correlation coefficients are TRUE?
Correct answers: All (A, B, C, D, E) are TRUE
Explanation:
A) TRUE – Pearson’s correlation assumes that both variables are approximately normally distributed.
B) TRUE – Spearman’s rank correlation is the non-parametric equivalent of Pearson’s. It uses ranks rather than raw values.
C) TRUE – Spearman’s correlation is resistant to outliers because it works with ranks, not actual values.
D) TRUE – Spearman’s correlation is appropriate when variables are ordinal (e.g., cancer stage, performance status).
E) TRUE – When parametric assumptions are met, Pearson’s correlation is generally preferred as it uses all available information in the data.
A researcher finds a correlation coefficient of r = 0.1 between two variables. Which of the following could explain this low value?
Correct answers: B, C, D, E are TRUE
Explanation:
A) FALSE – A perfect linear relationship would give r = +1 or r = -1, not r = 0.1.
B) TRUE – The correlation coefficient only measures linear association. A strong curved relationship (e.g., quadratic) could have r close to 0.
C) TRUE – Heterogeneous subgroups can mask the true correlation. Each subgroup might have a strong correlation, but when combined, the overall r can be low.
D) TRUE – Outliers can substantially distort the correlation coefficient, either inflating or deflating its value.
E) TRUE – A low r value may simply indicate that the variables are not linearly related.
The Pearson correlation coefficient between systolic blood pressure (mmHg) and age (years) in a sample of 30 women was r = 0.72 (p < 0.001). Hence r² = 0.52. Which statements are TRUE?
Correct answers: A, C, E are TRUE
Explanation:
A) TRUE – The p-value < 0.001 provides strong evidence against the null hypothesis of no linear association.
B) FALSE – r² = 0.52 means 52% (not 72%) of variability is explained. The value r² (coefficient of determination) represents explained variance, not r itself.
C) TRUE – If 52% is explained, then 48% (i.e., 1 - 0.52 = 0.48) remains unexplained.
D) FALSE – Correlation does not imply causation. We cannot conclude that age causes blood pressure to rise from observational data alone.
E) TRUE – The hypothesis test examines H₀: ρ = 0 (no linear association in the population).
In a simple linear regression model Y = a + bx, which of the following statements are TRUE?
Correct answers: A, B, C are TRUE
Explanation:
A) TRUE – By definition, the intercept is where the regression line crosses the Y-axis, i.e., when x = 0.
B) TRUE – The slope indicates how much Y changes on average when x increases by one unit.
C) TRUE – The term “regression coefficient” is commonly used for the slope (b) in simple linear regression.
D) FALSE – The slope can be positive, negative, or zero depending on the relationship between x and Y.
E) FALSE – The gradient is another term for the slope (b), not the intercept (a).
Which of the following statements about linear regression residuals are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Residual = observed value (y) - predicted value (Y).
B) TRUE – Ordinary least squares (OLS) finds the line that minimises the sum of squared residuals.
C) TRUE – A key assumption of linear regression is that residuals are normally distributed with mean = 0.
D) TRUE – R² (coefficient of determination) quantifies how much of the outcome variation is explained by the predictor(s).
E) FALSE – A high R² indicates good model fit, but this does not guarantee clinical utility. A model may explain variation statistically without being useful for prediction or decision-making.
Which of the following are assumptions of linear regression?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Linear regression assumes a linear relationship between x and Y.
B) TRUE – Observations should be independent of each other.
C) TRUE – Residuals (not the outcome variable itself) should be normally distributed.
D) TRUE – Homoscedasticity means the spread of residuals is constant across all fitted values.
E) FALSE – There is no assumption that the predictor (x) is normally distributed. The normality assumption applies to the residuals.
A researcher plots residuals against fitted values for a linear regression model. Which of the following would indicate a problem?
Correct answers: B, C indicate problems
Explanation:
A) NOT A PROBLEM – Random scatter around zero is the ideal pattern, indicating the linearity assumption is met.
B) PROBLEM – A curved pattern suggests a non-linear relationship, violating the linearity assumption.
C) PROBLEM – A funnel shape indicates heteroscedasticity (non-constant variance), violating the homoscedasticity assumption.
D) NOT A PROBLEM – Points falling on the Q-Q plot reference line support the normality assumption for residuals.
E) NOT A PROBLEM – A horizontal line at zero indicates no systematic pattern, supporting model assumptions.
Which of the following statements about multiple linear regression are TRUE?
Correct answers: All (A, B, C, D, E) are TRUE
Explanation:
A) TRUE – Multiple regression extends simple regression to include multiple predictors.
B) TRUE – Adjusted R² penalises for adding more predictors, preventing artificial inflation of R².
C) TRUE – For a binary predictor (e.g., treatment vs control), the coefficient represents the average difference in response between the two groups, holding other variables constant.
D) TRUE – The intercept is the predicted value of Y when all x values = 0.
E) TRUE – The same assumptions (linearity, independence, normality of residuals, homoscedasticity) apply.
Which of the following statements about logistic regression are TRUE?
Correct answers: A, B, C, E are TRUE
Explanation:
A) TRUE – Logistic regression is designed for binary outcomes (e.g., alive/dead, response/no response).
B) TRUE – The logistic function constrains predictions to the 0-1 probability range.
C) TRUE – Coefficients in logistic regression are often exponentiated to give odds ratios.
D) FALSE – Logistic regression uses a log-odds (logit) transformation, not the simple linear equation Y = a + bx.
E) TRUE – In R, glm(y ~ x, data = data, family = binomial) specifies logistic regression.
A study finds a strong positive correlation (r = 0.85) between ice cream sales and drowning deaths. Which statements are TRUE?
Correct answers: B, C, E are TRUE
Explanation:
A) FALSE – Correlation alone cannot prove causation.
B) TRUE – Hot weather likely increases both ice cream consumption and swimming, leading to more drownings. This is a classic confounding scenario.
C) TRUE – This fundamental principle means that observing an association does not establish a causal relationship.
D) FALSE – This would be inappropriate as ice cream is not the cause of drowning.
E) TRUE – A spurious correlation is one where two variables are associated but not causally related, often due to a confounding variable.
Which of the following statements about censoring in survival analysis are TRUE?
Correct answers: A, B, C, E are TRUE
Explanation:
A) TRUE – Censoring means the true survival time is unknown, only that it exceeds the observed follow-up time.
B) TRUE – These patients have not experienced the event by the study end; their true survival time is longer than observed.
C) TRUE – Loss to follow-up is a form of censoring as we do not know if/when the event occurred after the last observation.
D) FALSE – Ignoring censored patients wastes valuable information and can bias results. Proper survival methods account for censoring.
E) TRUE – The Kaplan-Meier estimator explicitly accounts for censored observations in its calculations.
Which of the following statements about the Kaplan-Meier method are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – The KM method recalculates survival probability at each event time.
B) TRUE – Cumulative survival = product of individual survival probabilities at each preceding event time.
C) TRUE – When a patient is censored, they are removed from the “at risk” denominator for subsequent time points.
D) TRUE – KM curves appear as step functions, with drops occurring at event times.
E) FALSE – KM assumes censoring is non-informative (i.e., unrelated to prognosis). If censoring is related to outcome, the estimates may be biased.
Which of the following statements about median survival are TRUE?
Correct answers: A, C, D, E are TRUE
Explanation:
A) TRUE – By definition, median survival is the time when S(t) = 0.50.
B) FALSE – If the curve does not drop to 0.50 (e.g., if follow-up is insufficient or too few events occurred), the median cannot be estimated.
C) TRUE – Survival data are typically skewed, making the median more robust than the mean.
D) TRUE – The median requires the survival function to reach 50%, which may not happen with limited follow-up.
E) TRUE – Graphically, the median is read from where the horizontal line at S(t) = 0.5 intersects the curve.
A clinical trial reports median overall survival of 12 months for treatment A and 14 months for treatment B. Which statements are TRUE?
Correct answers: B, C, D are TRUE
Explanation:
A) FALSE – “2 months longer” applies to the median, not the average (mean). Skewed distributions mean the mean could differ.
B) TRUE – By definition, median survival is when 50% of patients remain alive.
C) TRUE – Landmark survival (e.g., 1-year or 5-year survival) provides additional useful information about the survival curve.
D) TRUE – For skewed distributions (common in survival data), mean and median can differ substantially.
E) FALSE – The best summary depends on context. Landmark survival, mean restricted survival, or hazard ratios may be more informative in different situations.
Which of the following statements about the log-rank test are TRUE?
Correct answers: A, B, D are TRUE
Explanation:
A) TRUE – The log-rank test compares survival functions between groups.
B) TRUE – It tests H₀: survival curves are equal vs H₁: survival curves differ.
C) FALSE – The log-rank test only provides a p-value, not a measure of effect size like the hazard ratio.
D) TRUE – The test assumes hazards are proportional over time between groups.
E) FALSE – A significant p-value means we reject the null hypothesis that curves are equal (i.e., there is evidence of a difference).
Which of the following describes proportional hazards?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – This is the definition of proportional hazards: HR remains constant over the study duration.
B) TRUE – Crossing curves suggest the treatment effect changes direction over time, violating proportionality.
C) TRUE – The log-rank test assumes proportional hazards; violations can affect its validity.
D) TRUE – Schoenfeld residuals are a standard diagnostic tool for checking the proportional hazards assumption.
E) FALSE – Proportional hazards does not mean curves are parallel on the survival scale. On a log-cumulative hazard plot, parallel lines would indicate proportional hazards.
Which of the following statements about hazard ratios (HR) are TRUE?
Correct answers: A, B, C, E are TRUE
Explanation:
A) TRUE – HR = 0.5 means the instantaneous risk of the event is 50% lower in the treatment group.
B) TRUE – HR = 1.0 indicates equal hazards between groups.
C) TRUE – Cox regression is the standard method for estimating hazard ratios.
D) FALSE – HR relates to the rate of events, not the duration of survival. An HR of 2.0 means the event occurs at twice the rate, not that survival time is doubled.
E) TRUE – The HR interpretation assumes hazards remain proportional throughout follow-up.
Which of the following statements about Cox regression are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Cox regression is semi-parametric because it does not assume a specific distribution for the baseline hazard.
B) TRUE – The baseline hazard cancels out when calculating hazard ratios, so it does not need to be specified.
C) TRUE – Cox regression can include multiple predictors to adjust for confounding.
D) TRUE – Standard output includes HRs, 95% CIs, and p-values for each covariate.
E) FALSE – Cox regression handles censored observations appropriately; not all patients need to have had the event.
From a Kaplan-Meier curve, which of the following can be directly determined?
Correct answers: A, B, C, E are TRUE
Explanation:
A) TRUE – Median survival is read where the curve crosses 0.5 on the y-axis.
B) TRUE – Any landmark survival can be read from the y-axis at the corresponding x-axis time.
C) TRUE – Numbers at risk are typically displayed below the x-axis in published figures.
D) FALSE – Hazard ratios require Cox regression analysis; they cannot be read directly from KM curves.
E) TRUE – Censored observations are marked with tick marks or vertical lines on the curve.
A phase III trial reports HR = 0.75 (95% CI: 0.58-0.96, p = 0.02) for overall survival comparing new treatment vs standard. Which statements are TRUE?
Correct answers: A, B, D are TRUE
Explanation:
A) TRUE – HR = 0.75 means a 25% reduction in the instantaneous hazard of death (1 - 0.75 = 0.25).
B) TRUE – p = 0.02 < 0.05, so the result is statistically significant.
C) FALSE – The HR describes the rate of events, not survival duration. We cannot directly translate HR to percentage survival time improvement.
D) TRUE – If the 95% CI excludes 1.0 and lies entirely below 1.0 (for a beneficial treatment), the p-value will be < 0.05.
E) FALSE – The HR does not provide sufficient information to calculate median survival; the actual KM curves are needed.
In Kaplan-Meier analysis, the “number at risk” at each time point represents:
Correct answers: A, C are TRUE
Explanation:
A) TRUE – Number at risk includes patients still under observation and event-free.
B) FALSE – This would be the cumulative number of events, not the number at risk.
C) TRUE – The at-risk population forms the denominator for calculating survival probability at each event time.
D) FALSE – The at-risk number decreases over time as patients experience events or are censored.
E) FALSE – Censored patients are removed from the at-risk group; they are not included in the at-risk count after their censoring time.
Which of the following statements about sensitivity are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Sensitivity = proportion of truly diseased who are correctly identified by the test.
B) TRUE – Sensitivity = TP / (TP + FN) = a/(a+c).
C) TRUE – “SnNOut” = Sensitive test, Negative result, rules Out disease. A negative result on a highly sensitive test makes disease unlikely.
D) TRUE – The true positive rate is another name for sensitivity.
E) FALSE – Sensitivity is an intrinsic property of the test and does not change with prevalence. However, predictive values do depend on prevalence.
Which of the following statements about specificity are TRUE?
Correct answers: All (A, B, C, D, E) are TRUE
Explanation:
A) TRUE – Specificity = proportion of truly non-diseased who test negative.
B) TRUE – Specificity = TN / (TN + FP) = d/(b+d).
C) TRUE – “SpPIn” = Specific test, Positive result, rules In disease. A positive result on a highly specific test makes disease likely.
D) TRUE – True negative rate is synonymous with specificity.
E) TRUE – 100% specificity means all non-diseased individuals test negative, so there are no false positives.
Which of the following statements about positive predictive value (PPV) are TRUE?
Correct answers: A, B, C, E are TRUE
Explanation:
A) TRUE – PPV answers: “If the test is positive, what is the probability the patient has the disease?”
B) TRUE – PPV = TP / (TP + FP) = a/(a+b).
C) TRUE – Higher prevalence means more true positives relative to false positives, increasing PPV.
D) FALSE – Unlike sensitivity and specificity, PPV depends on disease prevalence.
E) TRUE – Low PPV means that among those who test positive, many do not actually have the disease.
Which of the following statements about negative predictive value (NPV) are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – NPV answers: “If the test is negative, what is the probability the patient does not have disease?”
B) TRUE – NPV = TN / (TN + FN) = d/(c+d).
C) TRUE – Higher prevalence means more diseased individuals, potentially more false negatives, reducing NPV.
D) TRUE – High NPV means a negative test result strongly suggests absence of disease.
E) FALSE – NPV and sensitivity measure different things. Sensitivity is a property of the test; NPV depends on prevalence.
A screening test has 90% sensitivity and 90% specificity. If applied to a population with 1% disease prevalence (N = 10,000), how many false positives would you expect?
Correct answer: C (approximately 990)
Explanation:
With 1% prevalence in 10,000 people:
With 90% specificity:
This illustrates that even with high specificity, screening low-prevalence conditions produces many false positives, resulting in low PPV.
Which of the following statements about likelihood ratios are TRUE?
Correct answers: A, B, D, E are TRUE
Explanation:
A) TRUE – LR+ = sensitivity / (1 - specificity) = sensitivity / false positive rate.
B) TRUE – LR+ indicates how much more likely a positive test is in those with disease versus without.
C) FALSE – The correct formula is LR- = (1 - sensitivity) / specificity.
D) TRUE – LRs depend only on sensitivity and specificity, not prevalence.
E) TRUE – LR = 1 means the test result is equally likely whether disease is present or absent.
Which of the following statements about accuracy and precision are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Accuracy (validity) is about closeness to the true value.
B) TRUE – Precision (reliability) is about reproducibility.
C) TRUE – Systematic error (bias) shifts all measurements in one direction, reducing accuracy while precision may remain high.
D) TRUE – Random error increases variability between measurements, reducing precision.
E) FALSE – Measurements can be precise (consistent) but inaccurate (biased). For example, a faulty scale might consistently give readings 2kg too high.
A test for prostate cancer has the following results: True positives = 167, False positives = 508, False negatives = 282, True negatives = 1993. Calculate the appropriate measures.
Which of the following are correct?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Sensitivity = 167/(167+282) = 167/449 = 0.37 = 37%
B) TRUE – Specificity = 1993/(508+1993) = 1993/2501 = 0.80 = 80%
C) TRUE – PPV = 167/(167+508) = 167/675 = 0.25 = 25%
D) TRUE – NPV = 1993/(282+1993) = 1993/2275 = 0.88 = 88%
E) FALSE – Prevalence = (167+282)/2950 = 449/2950 = 0.15 = 15% (This is actually true! E is correct.)
Note: All statements including E are actually correct. Prevalence = total diseased/total population = 449/2950 ≈ 15%.
Which of the following statements about screening versus diagnostic tests are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Screening is for apparently healthy individuals at risk.
B) TRUE – Diagnostic tests are used to confirm or exclude disease after a positive screen or clinical suspicion.
C) TRUE – High sensitivity minimises false negatives (missed cases) in screening.
D) TRUE – High specificity minimises false positives in diagnostic testing, avoiding unnecessary treatment.
E) FALSE – Positive screening tests indicate suspicion of disease and require confirmatory diagnostic testing.
Match the screening test with the cancer. Which statements are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Mammography is the primary breast cancer screening method in Scotland (ages 50-70, every 3 years).
B) TRUE – HPV testing has replaced cytology as the primary cervical screening test (ages 25-64, every 5 years).
C) TRUE – FIT has replaced the older guaiac-based test for bowel cancer screening (ages 50-74, every 2 years).
D) TRUE – Scottish bowel screening is offered biennially.
E) FALSE – Colposcopy is a confirmatory diagnostic test, not the primary screening test. HPV testing is the primary screen.
Using the standard 2×2 table layout (rows: test result +/-; columns: disease +/-), which cells represent:
Correct answers: All (A, B, C, D, E) are TRUE
Explanation:
The standard 2×2 table layout:
| Disease + | Disease - | |
|---|---|---|
| Test + | a (TP) | b (FP) |
| Test - | c (FN) | d (TN) |
Which of the following statements about clinical trial phases are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Phase I trials establish safe dosing, often using dose-escalation designs like “3+3”.
B) TRUE – Phase II trials screen for activity signals in specific cancer types.
C) TRUE – Phase III trials are confirmatory, comparing the new treatment to the current standard.
D) TRUE – Phase IV (post-marketing) studies evaluate real-world effectiveness and long-term safety.
E) FALSE – Phase I trials typically enrol small numbers of patients (often 20-50). Phase III trials enrol hundreds to thousands.
Which of the following statements about randomisation are TRUE?
Correct answers: A, C, D, E are TRUE
Explanation:
A) TRUE – Randomisation’s key advantage is balancing both measured and unmeasured confounders.
B) FALSE – Statistical analysis is still required to quantify treatment effects and uncertainty.
C) TRUE – Block randomisation ensures groups remain approximately equal throughout recruitment.
D) TRUE – Stratification by factors like performance status ensures balance on important prognostics.
E) TRUE – Random allocation prevents investigators from selectively assigning patients.
Which of the following statements about blinding are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Single-blind typically means patients are unaware of their allocation.
B) TRUE – Double-blind means both patients and those assessing outcomes are unaware.
C) TRUE – Blinding prevents behaviour changes (performance bias) and biased outcome assessment (detection bias).
D) TRUE – Open-label trials have no blinding; all parties know allocations.
E) FALSE – Blinding is not always possible, especially when treatments have distinct side effects (e.g., chemotherapy vs surgery) or routes of administration.
Which of the following correctly describes trial design types?
Correct answers: All (A, B, C, D, E) are TRUE
Explanation:
A) TRUE – In parallel designs, patients remain on their assigned treatment throughout.
B) TRUE – Crossover designs have patients act as their own controls by receiving treatments in different periods.
C) TRUE – Factorial designs (e.g., 2×2) test combinations of treatments efficiently.
D) TRUE – Superiority trials test whether the new treatment is statistically better.
E) TRUE – Non-inferiority trials test whether the new treatment is not worse than control by a specified margin.
In the PICO framework for clinical trials, what do the letters represent?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – P defines the patient population and eligibility criteria.
B) TRUE – I is the intervention or treatment being evaluated.
C) TRUE – C is the comparator (often standard of care or placebo).
D) TRUE – O specifies the primary and secondary outcomes.
E) FALSE – O stands for Outcome, not Odds ratio.
Which of the following are valid endpoints in oncology trials?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – OS (time from randomisation to death from any cause) is the gold standard endpoint.
B) TRUE – PFS (time to progression or death) is commonly used as a surrogate for OS.
C) TRUE – ORR (proportion achieving complete or partial response) measures tumour shrinkage.
D) TRUE – Quality of life is an important patient-centred outcome.
E) FALSE – Patient preference is not typically used as a clinical trial efficacy endpoint, though it may inform treatment decisions.
Which of the following statements about statistical errors in trials are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Type I error = rejecting a true null hypothesis = false positive.
B) TRUE – Type II error = failing to reject a false null hypothesis = false negative.
C) TRUE – α = 0.05 is the conventional threshold for statistical significance.
D) TRUE – Power = probability of correctly detecting a true effect = 1 - β.
E) FALSE – Increasing α (accepting higher false positive risk) actually decreases required sample size. Decreasing α increases sample size.
Which factors affect the required sample size for a clinical trial?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – Smaller α (stricter significance threshold) requires larger sample sizes.
B) TRUE – Higher power (e.g., 90% vs 80%) requires more patients.
C) TRUE – Detecting smaller differences requires larger samples.
D) TRUE – Greater outcome variability requires larger samples to detect the same effect.
E) FALSE – Capsule colour has no statistical bearing on sample size calculations!
Which of the following statements about intention-to-treat (ITT) analysis are TRUE?
Correct answers: A, C, D, E are TRUE
Explanation:
A) TRUE – ITT includes all randomised patients in their original groups regardless of compliance.
B) FALSE – ITT keeps all patients in their assigned groups even if they crossed over.
C) TRUE – ITT maintains balanced groups created by randomisation.
D) TRUE – ITT may dilute treatment effects if many patients switch treatments, giving conservative estimates.
E) TRUE – Per-protocol analysis excludes protocol violators and may overestimate treatment effects.
Which of the following statements about non-inferiority trials are TRUE?
Correct answers: A, B, C, D are TRUE
Explanation:
A) TRUE – The null hypothesis is that the new treatment is inferior by at least the margin.
B) TRUE – The margin must be pre-specified based on clinical judgement and historical data.
C) TRUE – Non-inferiority is demonstrated when the entire 95% CI lies above (for benefits) or below (for harms) the margin.
D) TRUE – Non-inferiority designs are justified when the new treatment offers other benefits (e.g., fewer side effects, oral administration).
E) FALSE – If the 95% CI also excludes 1.0 (no difference), superiority can be claimed in addition to non-inferiority.
Which of the following statements about interim analyses are TRUE?
Correct answers: All (A, B, C, D, E) are TRUE
Explanation:
A) TRUE – Trials can stop early if results are clear, saving resources and getting effective treatments to patients faster.
B) TRUE – Each look at the data increases the chance of a false positive finding.
C) TRUE – Methods like O’Brien-Fleming allocate α across interim and final analyses to preserve the overall 5% error rate.
D) TRUE – The number and timing of interim analyses should be pre-planned.
E) TRUE – Independent DSMBs review unblinded interim data to protect patient safety and trial integrity.
Arrange the following study types from highest to lowest level of evidence:
Which ordering is correct?
Correct ordering: A > B > C > D > E
Explanation:
The hierarchy of evidence (highest to lowest):
Note: This hierarchy applies to questions about treatment effectiveness. For prognosis or diagnosis, different designs may be more appropriate.
Which of the following are limitations of using historical controls in clinical trials?
Correct answers: All (A, B, C, D, E) are TRUE
Explanation:
A) TRUE – Treatment standards evolve, making historical comparisons problematic.
B) TRUE – Eligibility criteria may differ between the historical cohort and current trial.
C) TRUE – Better supportive care improves outcomes independently of the experimental treatment.
D) TRUE – Without randomisation, systematic differences between groups may exist.
E) TRUE – Earlier detection or stage migration can make current patients appear to do better even without treatment benefit.
These questions are designed for FRCR Part 1 and SCE Medical Oncology examination preparation. Based on course materials from Edinburgh Cancer Informatics.