onco-stats

Day 2 Revision Questions

These questions cover the Day 2 topics: Correlation and Regression, Survival Analysis, Screening Tests, and Clinical Trial Design.


Section 1: Correlation and Regression


Question 1.1: Correlation Coefficient Basics

Which of the following statements about the correlation coefficient are TRUE?

Click to reveal answer

Correct answers: A, C, D, E are TRUE

Explanation:


Question 1.2: Pearson vs Spearman Correlation

Which of the following statements about Pearson’s and Spearman’s correlation coefficients are TRUE?

Click to reveal answer

Correct answers: All (A, B, C, D, E) are TRUE

Explanation:


Question 1.3: When Not to Calculate Correlation

A researcher finds a correlation coefficient of r = 0.1 between two variables. Which of the following could explain this low value?

Click to reveal answer

Correct answers: B, C, D, E are TRUE

Explanation:


Question 1.4: Interpretation of Correlation

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?

Click to reveal answer

Correct answers: A, C, E are TRUE

Explanation:


Question 1.5: Linear Regression Basics

In a simple linear regression model Y = a + bx, which of the following statements are TRUE?

Click to reveal answer

Correct answers: A, B, C are TRUE

Explanation:


Question 1.6: Residuals and Model Fit

Which of the following statements about linear regression residuals are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 1.7: Assumptions of Linear Regression

Which of the following are assumptions of linear regression?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 1.8: Checking Model Assumptions

A researcher plots residuals against fitted values for a linear regression model. Which of the following would indicate a problem?

Click to reveal answer

Correct answers: B, C indicate problems

Explanation:


Question 1.9: Multiple Linear Regression

Which of the following statements about multiple linear regression are TRUE?

Click to reveal answer

Correct answers: All (A, B, C, D, E) are TRUE

Explanation:


Question 1.10: Logistic Regression

Which of the following statements about logistic regression are TRUE?

Click to reveal answer

Correct answers: A, B, C, E are TRUE

Explanation:


Question 1.11: Correlation vs Causation

A study finds a strong positive correlation (r = 0.85) between ice cream sales and drowning deaths. Which statements are TRUE?

Click to reveal answer

Correct answers: B, C, E are TRUE

Explanation:


Section 2: Survival Analysis


Question 2.1: Censoring

Which of the following statements about censoring in survival analysis are TRUE?

Click to reveal answer

Correct answers: A, B, C, E are TRUE

Explanation:


Question 2.2: Kaplan-Meier Estimator

Which of the following statements about the Kaplan-Meier method are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 2.3: Median Survival

Which of the following statements about median survival are TRUE?

Click to reveal answer

Correct answers: A, C, D, E are TRUE

Explanation:


Question 2.4: Summary Measures in Survival

A clinical trial reports median overall survival of 12 months for treatment A and 14 months for treatment B. Which statements are TRUE?

Click to reveal answer

Correct answers: B, C, D are TRUE

Explanation:


Question 2.5: Log-Rank Test

Which of the following statements about the log-rank test are TRUE?

Click to reveal answer

Correct answers: A, B, D are TRUE

Explanation:


Question 2.6: Proportional Hazards

Which of the following describes proportional hazards?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 2.7: Hazard Ratio

Which of the following statements about hazard ratios (HR) are TRUE?

Click to reveal answer

Correct answers: A, B, C, E are TRUE

Explanation:


Question 2.8: Cox Proportional Hazards Regression

Which of the following statements about Cox regression are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 2.9: Reading Kaplan-Meier Curves

From a Kaplan-Meier curve, which of the following can be directly determined?

Click to reveal answer

Correct answers: A, B, C, E are TRUE

Explanation:


Question 2.10: Survival Analysis in Clinical Trials

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?

Click to reveal answer

Correct answers: A, B, D are TRUE

Explanation:


Question 2.11: Number at Risk

In Kaplan-Meier analysis, the “number at risk” at each time point represents:

Click to reveal answer

Correct answers: A, C are TRUE

Explanation:


Section 3: Screening Tests


Question 3.1: Sensitivity

Which of the following statements about sensitivity are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 3.2: Specificity

Which of the following statements about specificity are TRUE?

Click to reveal answer

Correct answers: All (A, B, C, D, E) are TRUE

Explanation:


Question 3.3: Positive Predictive Value

Which of the following statements about positive predictive value (PPV) are TRUE?

Click to reveal answer

Correct answers: A, B, C, E are TRUE

Explanation:


Question 3.4: Negative Predictive Value

Which of the following statements about negative predictive value (NPV) are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 3.5: Effect of 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?

Click to reveal answer

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.


Question 3.6: Likelihood Ratios

Which of the following statements about likelihood ratios are TRUE?

Click to reveal answer

Correct answers: A, B, D, E are TRUE

Explanation:


Question 3.7: Accuracy vs Precision

Which of the following statements about accuracy and precision are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 3.8: Screening Test Properties – Calculation

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?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:

Note: All statements including E are actually correct. Prevalence = total diseased/total population = 449/2950 ≈ 15%.


Question 3.9: Screening vs Diagnostic Tests

Which of the following statements about screening versus diagnostic tests are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 3.10: Scottish Screening Programmes

Match the screening test with the cancer. Which statements are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 3.11: 2×2 Table Interpretation

Using the standard 2×2 table layout (rows: test result +/-; columns: disease +/-), which cells represent:

Click to reveal answer

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)

Section 4: Clinical Trial Design


Question 4.1: Phases of Clinical Trials

Which of the following statements about clinical trial phases are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 4.2: Randomisation

Which of the following statements about randomisation are TRUE?

Click to reveal answer

Correct answers: A, C, D, E are TRUE

Explanation:


Question 4.3: Blinding

Which of the following statements about blinding are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 4.4: Trial Design Types

Which of the following correctly describes trial design types?

Click to reveal answer

Correct answers: All (A, B, C, D, E) are TRUE

Explanation:


Question 4.5: PICO Framework

In the PICO framework for clinical trials, what do the letters represent?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 4.6: Trial Endpoints

Which of the following are valid endpoints in oncology trials?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 4.7: Type I and Type II Errors

Which of the following statements about statistical errors in trials are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 4.8: Sample Size Considerations

Which factors affect the required sample size for a clinical trial?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 4.9: Intention-to-Treat Analysis

Which of the following statements about intention-to-treat (ITT) analysis are TRUE?

Click to reveal answer

Correct answers: A, C, D, E are TRUE

Explanation:


Question 4.10: Non-Inferiority Trials

Which of the following statements about non-inferiority trials are TRUE?

Click to reveal answer

Correct answers: A, B, C, D are TRUE

Explanation:


Question 4.11: Interim Analyses

Which of the following statements about interim analyses are TRUE?

Click to reveal answer

Correct answers: All (A, B, C, D, E) are TRUE

Explanation:


Question 4.12: Evidence Hierarchy

Arrange the following study types from highest to lowest level of evidence:

Which ordering is correct?

Click to reveal answer

Correct ordering: A > B > C > D > E

Explanation:

The hierarchy of evidence (highest to lowest):

  1. Systematic reviews and meta-analyses of RCTs – Synthesise multiple high-quality trials
  2. Well-designed RCTs – Gold standard for testing interventions with randomisation
  3. Cohort studies – Observational but follow groups over time
  4. Case-control studies – Compare those with/without disease retrospectively
  5. Expert opinion – Lowest level; subjective and prone to bias

Note: This hierarchy applies to questions about treatment effectiveness. For prognosis or diagnosis, different designs may be more appropriate.


Question 4.13: Historical Controls

Which of the following are limitations of using historical controls in clinical trials?

Click to reveal answer

Correct answers: All (A, B, C, D, E) are TRUE

Explanation:


These questions are designed for FRCR Part 1 and SCE Medical Oncology examination preparation. Based on course materials from Edinburgh Cancer Informatics.