FRCR ONCOLOGY PART 1 - ALL FREE PRACTICE MCQS: FRCR Oncology Part 1: MEDICAL STATISTICS

What type of variable is “tumour grade” (I, II, III)?
A. Nominal
B. Ordinal
C. Interval
D. Ratio
E. Binary
A study reports a p-value of 0.03. What does this mean?
A. The null hypothesis is false
B. The alternative hypothesis is true
C. There is a 3% chance the result is due to random variation under the null hypothesis
D. There is a 97% chance the result is clinically significant
E. The study is underpowered
In a normal distribution, approximately what percentage of values lie within ±2 standard deviations of the mean?
A. 50%
B. 68%
C. 75%
D. 95%
E. 99%
What is the most appropriate measure of central tendency for a skewed dataset?
A. Mean
B. Mode
C. Standard deviation
D. Median
E. Interquartile range
What statistical test compares means between more than two groups?
A. Chi-square
B. Student’s t-test
C. ANOVA
D. Fisher’s exact test
E. Mann-Whitney U test
What does the 95% confidence interval represent?
A. There is a 95% chance the result is not due to chance
B. In 95% of samples, the interval will contain the true population parameter
C. There is a 95% probability the mean lies within the interval
D. It shows the range of all possible values
E. It indicates statistical significance
The null hypothesis in a clinical trial usually states:
A. There is a difference between groups
B. The treatment is effective
C. The treatment is superior
D. There is no difference between groups
E. The results are not statistically significant
Which of the following reduces the likelihood of a Type I error?
A. Increasing power
B. Increasing sample size
C. Decreasing alpha level
D. Increasing beta
E. Using one-tailed test
What is the primary purpose of randomisation in clinical trials?
A. Improve generalisability
B. Eliminate confounding
C. Reduce measurement error
D. Increase statistical power
E. Achieve blinding
Which of the following is NOT a feature of a case-control study?
A. Retrospective in nature
B. Suitable for rare diseases
C. Cannot calculate incidence
D. Time-consuming and expensive
E. Prone to recall bias

What is the best statistical test to compare survival distributions between two groups?
A. Student’s t-test
B. Mann-Whitney U test
C. Chi-square test
D. Kaplan-Meier method
E. Log-rank test
Which term refers to the proportion of true positives correctly identified?
A. Specificity
B. Sensitivity
C. Positive predictive value
D. Negative predictive value
E. Accuracy
What statistical method adjusts for multiple covariates in survival analysis?
A. Linear regression
B. Logistic regression
C. Cox proportional hazards model
D. Kaplan-Meier method
E. Paired t-test
In a ROC curve, what does the area under the curve (AUC) represent?
A. Incidence
B. Risk ratio
C. Diagnostic accuracy
D. Statistical significance
E. Prevalence
What is an example of a continuous variable?
A. Tumour location
B. Tumour grade
C. Survival time
D. Gender
E. Histological type
What does a wide confidence interval imply?
A. Low variability
B. Precise estimate
C. Large sample size
D. High variability or low sample size
E. Statistical significance
A Type II error occurs when:
A. The null hypothesis is wrongly rejected
B. The null hypothesis is wrongly accepted
C. The sample size is too small
D. A significant p-value is observed
E. Alpha is set too low
In regression analysis, the R-squared value represents:
A. Probability of error
B. Variance in outcome explained by model
C. Risk of bias
D. Standard error
E. Correlation coefficient
Which trial design uses patients as their own control?
A. Randomised controlled trial
B. Case-control study
C. Crossover trial
D. Cohort study
E. Non-inferiority trial
What does ‘intention-to-treat’ analysis prevent?
A. Information bias
B. Selection bias
C. Attrition bias
D. Confounding
E. Performance bias
Which measure is most appropriate for skewed data?
A. Mean
B. Mode
C. Standard deviation
D. Median
E. Variance
In a 2x2 table, which test is preferred for small sample sizes?
A. ANOVA
B. Chi-square
C. Fisher’s exact test
D. Mann-Whitney U test
E. Paired t-test
What does “power” in a study represent?
A. Likelihood of detecting an effect when one exists
B. Ability to avoid Type I error
C. Sample size
D. Effect size
E. Alpha level
What does ‘p-hacking’ refer to?
A. Randomisation failure
B. Selective reporting of outcomes
C. Manipulating data to reach significance
D. Power calculation error
E. Violation of ethics
Which is NOT a limitation of meta-analysis?
A. Publication bias
B. Heterogeneity
C. Selection bias
D. Low statistical power
E. Quality of included studies
Which measure best describes the strength of association in a case-control study?
A. Relative risk
B. Hazard ratio
C. Risk difference
D. Odds ratio
E. Incidence rate
Which test would best compare medians between two unpaired groups?
A. Chi-square
B. Paired t-test
C. Mann-Whitney U
D. ANOVA
E. Fisher’s exact test
What is the main limitation of using mean and standard deviation for non-normal data?
A. They are difficult to compute
B. They are not affected by outliers
C. They misrepresent the data’s distribution
D. They work only for binary data
E. They are qualitative measures
What statistical method assesses the influence of multiple variables on a binary outcome?
A. Logistic regression
B. Cox regression
C. Poisson regression
D. Linear regression
E. ANOVA
When should a Kaplan-Meier curve be used?
A. When comparing proportions
B. When analyzing continuous variables
C. When estimating time-to-event data
D. When plotting histograms
E. When calculating odds ratios

Which statistical test is most appropriate to compare proportions between two groups?
A. Student’s t-test
B. ANOVA
C. Chi-square test
D. Linear regression
E. Pearson correlation
What is the most appropriate graphical display for a categorical variable?
A. Histogram
B. Box plot
C. Bar chart
D. Scatter plot
E. Line graph
Which of the following statements is true of a correlation coefficient (r)?
A. It ranges from -2 to +2
B. It indicates causality
C. It measures strength and direction of linear relationship
D. It is used for categorical data
E. It is affected only by outliers
What kind of variable is “disease-free survival”?
A. Nominal
B. Ordinal
C. Categorical
D. Continuous
E. Binary
If a confidence interval for a hazard ratio includes 1, the result is:
A. Statistically significant
B. Not statistically significant
C. Clinically significant
D. Strongly associated
E. Biased
What does the term “hazard” in hazard ratio refer to?
A. Total risk over a period
B. Relative risk
C. Instantaneous event rate
D. Incidence rate
E. Mortality rate
What is the effect of increasing sample size on a study?
A. Decreases power
B. Increases p-value
C. Reduces Type II error
D. Increases Type I error
E. Widens confidence intervals
What type of data is used in Cox regression?
A. Categorical
B. Continuous only
C. Binary only
D. Time-to-event data
E. Nominal
Which of the following measures is least affected by outliers?
A. Mean
B. Standard deviation
C. Variance
D. Median
E. Range
Which assumption is required for the Cox proportional hazards model?
A. Normally distributed errors
B. Equal means
C. Homoscedasticity
D. Proportional hazards over time
E. Random sampling
What statistical concept is used in multiple hypothesis testing correction?
A. Odds ratio
B. Confidence interval
C. Bonferroni correction
D. Linear regression
E. Risk difference
What is multicollinearity in regression analysis?
A. Association between dependent variables
B. Non-linearity of variables
C. Strong correlation between independent variables
D. Random variation in data
E. Missing data bias
Which value of p is conventionally considered statistically significant?
A. p < 0.2
B. p < 0.1
C. p < 0.05
D. p < 0.01
E. p > 0.05
The standard error of the mean (SEM) is:
A. A measure of variability of a population
B. The square of standard deviation
C. Standard deviation divided by the square root of sample size
D. Independent of sample size
E. Same as standard deviation
Which of the following best reduces random error?
A. Stratified randomisation
B. Increasing sample size
C. Matching
D. Blinding
E. Adjusting for confounders
What does the term “relative risk” compare?
A. Odds of exposure in diseased vs. non-diseased
B. Risk of outcome in exposed vs. unexposed
C. Mean difference between groups
D. Proportion of risk attributable to exposure
E. Disease frequency
Which study design is most prone to recall bias?
A. Cohort study
B. Randomised controlled trial
C. Cross-sectional study
D. Case-control study
E. Ecological study
What is the function of a forest plot in meta-analysis?
A. Estimate p-value
B. Display risk of bias
C. Show effect sizes and confidence intervals
D. Show regression line
E. Assess heterogeneity
What does “I² statistic” measure in meta-analyses?
A. Sample size
B. Mean difference
C. Proportion of variance due to heterogeneity
D. Confidence interval
E. Publication bias
What is the principal risk when multiple comparisons are performed without adjustment?
A. Type II error
B. Decreased power
C. Increased confidence
D. Type I error inflation
E. Loss of blinding
What is the primary purpose of blinding in trials?
A. Reduce selection bias
B. Reduce attrition
C. Reduce measurement bias
D. Reduce confounding
E. Reduce heterogeneity
What is the definition of prevalence?
A. Rate of new cases
B. Rate of exposed individuals
C. Proportion of population with disease at a given time
D. Probability of surviving disease
E. Number of patients lost to follow-up
What type of bias arises when sicker patients are excluded from study analysis?
A. Lead-time bias
B. Selection bias
C. Measurement bias
D. Attrition bias
E. Performance bias
What is the denominator in calculating incidence rate?
A. Number of exposed individuals
B. Number of deaths
C. Person-time at risk
D. Total population
E. Number lost to follow-up
What does “confounding” mean?
A. Misclassification of exposure
B. Systematic error
C. Distortion of the association due to another variable
D. Chance variation
E. Measurement bias
What test is used to assess agreement between two observers?
A. Pearson correlation
B. Logistic regression
C. Bland-Altman plot
D. Chi-square test
E. Cohen’s kappa
What is an ecological fallacy?
A. Wrong time horizon
B. Assuming group data apply to individuals
C. Using retrospective data
D. Loss to follow-up
E. Contamination between groups
Which plot is used to check for publication bias?
A. Kaplan-Meier plot
B. Forest plot
C. Funnel plot
D. ROC curve
E. QQ plot
In logistic regression, the dependent variable must be:
A. Ordinal
B. Continuous
C. Binary
D. Nominal
E. Interval
Which measure tells you the degree to which repeated measurements give consistent results?
A. Validity
B. Specificity
C. Accuracy
D. Precision
E. Bias

What does the term "effect size" refer to in clinical trials?
A. The p-value of the result
B. Magnitude of the observed difference between groups
C. Number of patients in the study
D. Width of the confidence interval
E. The type of statistical test used
What statistical method is appropriate for paired categorical data?
A. Chi-square test
B. McNemar’s test
C. Student’s t-test
D. Fisher’s exact test
E. Wilcoxon signed-rank test
The primary purpose of stratified randomisation is to:
A. Ensure blinding
B. Increase sample size
C. Balance known confounders
D. Reduce loss to follow-up
E. Increase statistical significance
What does a p-value of 0.25 mean?
A. Statistically significant result
B. 25% probability the result is not due to chance
C. Fail to reject the null hypothesis
D. Accept the alternative hypothesis
E. High power
What assumption must be met in linear regression?
A. Independent variable must be categorical
B. Residuals are not normally distributed
C. Homoscedasticity
D. Binary outcome
E. Variables are nominal
A funnel plot is symmetric. What does this suggest?
A. High heterogeneity
B. No publication bias
C. Large effect size
D. Poor quality studies
E. High p-values
What type of study is best for estimating incidence?
A. Case-control study
B. Cross-sectional study
C. Case series
D. Cohort study
E. Ecological study
Which is a method of dealing with confounding during analysis?
A. Randomisation
B. Stratification
C. Blinding
D. Concealment
E. Power calculation
What is the term for overestimating survival due to early diagnosis from screening?
A. Recall bias
B. Selection bias
C. Lead-time bias
D. Confounding
E. Attrition bias
What does “number needed to treat” (NNT) represent?
A. Number needed to detect a significant result
B. Number needed to cause harm
C. Number needed to prevent one outcome
D. Total sample size
E. Number of responders
A relative risk of 1 implies:
A. Strong association
B. Increased risk in the exposed group
C. No difference in risk
D. Reduced risk in exposed
E. Confounding is likely
What is the function of a Q-Q plot?
A. Compare means
B. Assess normality
C. Evaluate multicollinearity
D. Calculate p-values
E. Test heterogeneity
Which assumption is key in ANOVA?
A. Binary outcome
B. Independent observations
C. Ordinal predictors
D. Non-random sampling
E. Time-to-event data
What is true about a type I error?
A. Increases with sample size
B. Occurs when null is wrongly rejected
C. Indicates underpowering
D. Increases with beta
E. Is not related to significance
Which measure best reflects the reproducibility of a test?
A. Sensitivity
B. Specificity
C. Accuracy
D. Precision
E. Validity
What is the main limitation of observational studies in establishing causality?
A. Low sample size
B. High cost
C. Inability to control for bias/confounding
D. Unclear exposure
E. Limited external validity
In hypothesis testing, the beta value refers to:
A. Probability of type I error
B. Power
C. Effect size
D. Probability of type II error
E. Confidence interval
The interquartile range (IQR) represents:
A. Variability around the mean
B. Middle 50% of data
C. Standard deviation
D. Skewness
E. Outliers
The primary goal of a phase III clinical trial is to:
A. Determine pharmacokinetics
B. Assess safety in healthy volunteers
C. Compare effectiveness against standard treatment
D. Explore dose range
E. Monitor long-term outcomes
A small p-value means:
A. Large effect size
B. High power
C. Evidence against the null hypothesis
D. Bias is likely
E. Confidence intervals are wide
What is “surveillance bias”?
A. Systematic underreporting
B. Subjects monitored more closely are more likely to have outcomes detected
C. Lack of follow-up
D. Misclassification
E. Censoring error
Which test assesses whether two continuous variables are linearly related?
A. ANOVA
B. Spearman rank
C. Pearson correlation
D. Chi-square
E. Kaplan-Meier
A study finds a hazard ratio of 1.5 (95% CI 1.1–2.3). This means:
A. The result is not statistically significant
B. Hazard in treatment group is 50% higher, and CI excludes 1
C. The study is underpowered
D. Effect size is negligible
E. Random error occurred
In a 2x2 diagnostic test table, which is calculated as TP / (TP + FP)?
A. Sensitivity
B. Specificity
C. Negative predictive value
D. Positive predictive value
E. Accuracy
The primary purpose of allocation concealment is to:
A. Prevent performance bias
B. Prevent detection bias
C. Prevent selection bias
D. Prevent attrition bias
E. Ensure blinding
What is the role of a Data Monitoring Committee (DMC)?
A. Recruit patients
B. Publish trial outcomes
C. Monitor safety and efficacy during a trial
D. Fund the trial
E. Determine sample size
What does the “null value” refer to in a confidence interval?
A. The lowest value
B. A non-significant point estimate
C. The value indicating no effect (e.g., 1 for RR, 0 for mean difference)
D. The true population mean
E. The midpoint of the CI
Which of these is NOT a continuous variable?
A. Age
B. Blood pressure
C. Tumour size
D. Gender
E. Time to recurrence
Which of the following is used to compare two related quantitative samples?
A. Student’s t-test
B. Paired t-test
C. Chi-square test
D. ANOVA
E. Cox regression
What does "external validity" refer to?
A. Accuracy of outcome measure
B. Reproducibility of findings
C. Generalisability to other populations
D. Blinding effectiveness
E. Correctness of hypothesis
What is a Kaplan-Meier curve used to represent?
A. Risk ratios
B. Confidence intervals
C. Survival probability over time
D. Correlation
E. Prevalence
What is censoring in survival analysis?
A. Omission of data
B. Data cleaning
C. When the event has not occurred at last follow-up
D. Removal of outliers
E. Statistical significance
Which term describes variability of sample mean from population mean?
A. Bias
B. Standard deviation
C. Standard error
D. Confidence interval
E. Effect size
In diagnostic testing, what is specificity?
A. TP / (TP + FN)
B. TN / (TN + FP)
C. TP / (TP + FP)
D. TN / (TN + FN)
E. (TP + TN) / Total
What does heterogeneity in meta-analysis suggest?
A. All studies had same effect size
B. High statistical power
C. Variability in study results
D. Uniform population
E. Low external validity
In a standard normal distribution, what is the Z-score for the mean?
A. 0
B. 1
C. -1
D. 2
E. 100
What is the best design to study rare exposures?
A. Case-control
B. Cross-sectional
C. Cohort
D. Randomised trial
E. Ecological
A trial reports a risk difference of 5% (95% CI: -1% to 11%). What is the interpretation?
A. Statistically significant
B. Clinically meaningless
C. Could be no effect
D. Requires logistic regression
E. Underpowered
What is the benefit of using a likelihood ratio in diagnostics?
A. Easy to calculate
B. Does not depend on disease prevalence
C. Highly sensitive
D. Measures survival
E. Prevents bias
What is the main purpose of a Phase I clinical trial?
A. Compare treatment to placebo
B. Assess long-term outcomes
C. Determine maximum tolerated dose and safety
D. Evaluate efficacy
E. Assess blinding success

FRCR ONCOLOGY PART 1 - ALL FREE PRACTICE MCQS

Sunday, May 18, 2025

FRCR Oncology Part 1: MEDICAL STATISTICS - 9 (100 QUESTIONS, ANSWERS BELOW)

ANSWERS