PRACTICE FINAL VERIFIED EXAM | 2026/2027 UPDATE | PUBH 6002
BIOSTATISTICS FOR
PUBLIC HEALTH
PUBH 6002 - PRACTICE FINAL
Practice Final Verified Exam | 2026/2027 UPDATE | Comprehensive Review
2026/2027 UPDATE VERIFIED Q&A PRACTICE FINAL BIOSTATS
, EXAM COVERAGE AREAS
Biostatistics for Public Health - Practice Final Domains | 2026/2027 Update
Descriptive Statistics & Data Types
• Variable types: nominal, ordinal, interval, ratio
• Measures central tendency: mean, median, mode - when to use
• Measures dispersion: variance, SD, range, IQR
• Graphical displays: histogram, boxplot, scatterplot
Probability, Distributions & Sampling
• Probability rules, conditional probability, Bayes
• Normal distribution: z-scores, 68-95-99.7 rule
• Binomial, Poisson distributions
• Sampling methods, sampling distribution, CLT
Hypothesis Testing & Confidence Intervals
• Null vs alternative hypotheses, Type I/II error, power
• p-value interpretation, alpha 0.05
• Confidence intervals: interpretation, relation to hypothesis test
• One-sample, two-sample tests
T-Tests, ANOVA & Nonparametric Tests
• One-sample t-test, independent t-test, paired t-test - assumptions
• ANOVA: one-way, post-hoc tests, assumptions
• Chi-square test: goodness-of-fit, independence
• Nonparametric: Mann-Whitney, Wilcoxon, Kruskal-Wallis when to use
Correlation, Regression & Epidemiologic Measures
• Correlation vs causation, Pearson r, Spearman
• Simple linear regression: equation, R², assumptions
• Logistic regression: OR interpretation
• Epi measures: prevalence, incidence, RR, OR, sensitivity, specificity
Study Design & 2026/2027 Application
• Study designs: RCT, cohort, case-control, cross-sectional
• Bias, confounding, effect modification
• Sample size calculation basics
• Public health application: interpreting biostats literature
PubH 6002 Biostatistics for Public Health | Practice Final | 2026/2027 UPDATE Page 2
, QUESTION 1: PubH 6002: Which measure of central tendency best for skewed income
data?
A) Median - resistant to outliers, better for skewed
B) Mean always best
C) Mode always best
D) Range
CORRECT ANSWER: A) Median - resistant to outliers, better for skewed
RATIONALE: Mean affected by outliers/skewness, median resistant and better for skewed data like income.
Mode for categorical. For symmetric data mean=median=mode. Skewed right mean > median.
QUESTION 2: 68-95-99.7 rule for normal distribution means?
A) 68% within 1 SD, 95% within 2 SD, 99.7% within 3 SD of mean
B) 50% within 1 SD
C) 100% within 1 SD
D) No rule
CORRECT ANSWER: A) 68% within 1 SD, 95% within 2 SD, 99.7% within 3 SD of mean
RATIONALE: Empirical rule for normal distribution: ~68% observations within mean ±1 SD, ~95% within ±2
SD, ~99.7% within ±3 SD. Used for z-scores and probability.
QUESTION 3: p-value 0.03 with alpha 0.05 interpretation?
A) Reject null, statistically significant, 3% probability observing result if null true
B) Accept null
C) No significance
D) p-value is probability null true
CORRECT ANSWER: A) Reject null, statistically significant, 3% probability observing
result if null true
RATIONALE: p-value 0.03 < alpha 0.05 => reject null, statistically significant. p-value NOT probability null
true, but probability of observing data as extreme if null true. 0.03 means 3% chance if null true.
QUESTION 4: 95% CI for mean is 10 to 20 interpretation?
A) We are 95% confident true population mean lies between 10 and 20; if repeated 100 samples, 95
CIs would contain true mean
B) 95% of data between 10-20
C) Mean is 15 always
D) No interpretation
CORRECT ANSWER: A) We are 95% confident true population mean lies between 10
and 20; if repeated 100 samples, 95 CIs would contain true mean
RATIONALE: CI interpretation: 95% confident true population parameter in interval. If null value outside CI
=> reject null at alpha 0.05. CI width depends on SD, n, confidence level.
QUESTION 5: When to use independent t-test vs paired t-test?
A) Independent: compare means of two unrelated groups; Paired: same subjects before-after or
matched pairs
PubH 6002 Biostatistics for Public Health | Practice Final | 2026/2027 UPDATE Page 3
, B) Same test always
C) Paired for unrelated groups
D) Independent for same subjects
CORRECT ANSWER: A) Independent: compare means of two unrelated groups; Paired:
same subjects before-after or matched pairs
RATIONALE: Independent t-test: two independent groups (treatment vs control). Paired t-test: same
subjects measured twice (pre-post) or matched pairs. Paired more powerful as controls for subject
variability.
QUESTION 6: Chi-square test is used for?
A) Categorical data association - independence or goodness-of-fit
B) Continuous data means
C) Correlation
D) Regression
CORRECT ANSWER: A) Categorical data association - independence or goodness-of-fit
RATIONALE: Chi-square: test association between two categorical variables (independence) or compare
observed vs expected frequencies (goodness-of-fit). Assumptions: expected count >=5 in 80% cells.
QUESTION 7: Relative Risk vs Odds Ratio?
A) RR = incidence exposed / incidence unexposed - cohort/RCT; OR = odds exposed / odds
unexposed - case-control, approximates RR when disease rare
B) Same thing always
C) OR only for cohort
D) RR only for case-control
CORRECT ANSWER: A) RR = incidence exposed / incidence unexposed - cohort/RCT;
OR = odds exposed / odds unexposed - case-control, approximates RR when disease
rare
RATIONALE: RR from cohort/RCT: risk ratio. OR from case-control: odds ratio, approximates RR when
disease rare (<10%). RR=1 no association, >1 risk, <1 protective.
QUESTION 8: PubH 6002: Which measure of central tendency best for skewed income
data?
A) Median - resistant to outliers, better for skewed
B) Mean always best
C) Mode always best
D) Range
CORRECT ANSWER: A) Median - resistant to outliers, better for skewed
RATIONALE: Mean affected by outliers/skewness, median resistant and better for skewed data like income.
Mode for categorical. For symmetric data mean=median=mode. Skewed right mean > median.
QUESTION 9: 68-95-99.7 rule for normal distribution means?
A) 68% within 1 SD, 95% within 2 SD, 99.7% within 3 SD of mean
B) 50% within 1 SD
PubH 6002 Biostatistics for Public Health | Practice Final | 2026/2027 UPDATE Page 4
BIOSTATISTICS FOR
PUBLIC HEALTH
PUBH 6002 - PRACTICE FINAL
Practice Final Verified Exam | 2026/2027 UPDATE | Comprehensive Review
2026/2027 UPDATE VERIFIED Q&A PRACTICE FINAL BIOSTATS
, EXAM COVERAGE AREAS
Biostatistics for Public Health - Practice Final Domains | 2026/2027 Update
Descriptive Statistics & Data Types
• Variable types: nominal, ordinal, interval, ratio
• Measures central tendency: mean, median, mode - when to use
• Measures dispersion: variance, SD, range, IQR
• Graphical displays: histogram, boxplot, scatterplot
Probability, Distributions & Sampling
• Probability rules, conditional probability, Bayes
• Normal distribution: z-scores, 68-95-99.7 rule
• Binomial, Poisson distributions
• Sampling methods, sampling distribution, CLT
Hypothesis Testing & Confidence Intervals
• Null vs alternative hypotheses, Type I/II error, power
• p-value interpretation, alpha 0.05
• Confidence intervals: interpretation, relation to hypothesis test
• One-sample, two-sample tests
T-Tests, ANOVA & Nonparametric Tests
• One-sample t-test, independent t-test, paired t-test - assumptions
• ANOVA: one-way, post-hoc tests, assumptions
• Chi-square test: goodness-of-fit, independence
• Nonparametric: Mann-Whitney, Wilcoxon, Kruskal-Wallis when to use
Correlation, Regression & Epidemiologic Measures
• Correlation vs causation, Pearson r, Spearman
• Simple linear regression: equation, R², assumptions
• Logistic regression: OR interpretation
• Epi measures: prevalence, incidence, RR, OR, sensitivity, specificity
Study Design & 2026/2027 Application
• Study designs: RCT, cohort, case-control, cross-sectional
• Bias, confounding, effect modification
• Sample size calculation basics
• Public health application: interpreting biostats literature
PubH 6002 Biostatistics for Public Health | Practice Final | 2026/2027 UPDATE Page 2
, QUESTION 1: PubH 6002: Which measure of central tendency best for skewed income
data?
A) Median - resistant to outliers, better for skewed
B) Mean always best
C) Mode always best
D) Range
CORRECT ANSWER: A) Median - resistant to outliers, better for skewed
RATIONALE: Mean affected by outliers/skewness, median resistant and better for skewed data like income.
Mode for categorical. For symmetric data mean=median=mode. Skewed right mean > median.
QUESTION 2: 68-95-99.7 rule for normal distribution means?
A) 68% within 1 SD, 95% within 2 SD, 99.7% within 3 SD of mean
B) 50% within 1 SD
C) 100% within 1 SD
D) No rule
CORRECT ANSWER: A) 68% within 1 SD, 95% within 2 SD, 99.7% within 3 SD of mean
RATIONALE: Empirical rule for normal distribution: ~68% observations within mean ±1 SD, ~95% within ±2
SD, ~99.7% within ±3 SD. Used for z-scores and probability.
QUESTION 3: p-value 0.03 with alpha 0.05 interpretation?
A) Reject null, statistically significant, 3% probability observing result if null true
B) Accept null
C) No significance
D) p-value is probability null true
CORRECT ANSWER: A) Reject null, statistically significant, 3% probability observing
result if null true
RATIONALE: p-value 0.03 < alpha 0.05 => reject null, statistically significant. p-value NOT probability null
true, but probability of observing data as extreme if null true. 0.03 means 3% chance if null true.
QUESTION 4: 95% CI for mean is 10 to 20 interpretation?
A) We are 95% confident true population mean lies between 10 and 20; if repeated 100 samples, 95
CIs would contain true mean
B) 95% of data between 10-20
C) Mean is 15 always
D) No interpretation
CORRECT ANSWER: A) We are 95% confident true population mean lies between 10
and 20; if repeated 100 samples, 95 CIs would contain true mean
RATIONALE: CI interpretation: 95% confident true population parameter in interval. If null value outside CI
=> reject null at alpha 0.05. CI width depends on SD, n, confidence level.
QUESTION 5: When to use independent t-test vs paired t-test?
A) Independent: compare means of two unrelated groups; Paired: same subjects before-after or
matched pairs
PubH 6002 Biostatistics for Public Health | Practice Final | 2026/2027 UPDATE Page 3
, B) Same test always
C) Paired for unrelated groups
D) Independent for same subjects
CORRECT ANSWER: A) Independent: compare means of two unrelated groups; Paired:
same subjects before-after or matched pairs
RATIONALE: Independent t-test: two independent groups (treatment vs control). Paired t-test: same
subjects measured twice (pre-post) or matched pairs. Paired more powerful as controls for subject
variability.
QUESTION 6: Chi-square test is used for?
A) Categorical data association - independence or goodness-of-fit
B) Continuous data means
C) Correlation
D) Regression
CORRECT ANSWER: A) Categorical data association - independence or goodness-of-fit
RATIONALE: Chi-square: test association between two categorical variables (independence) or compare
observed vs expected frequencies (goodness-of-fit). Assumptions: expected count >=5 in 80% cells.
QUESTION 7: Relative Risk vs Odds Ratio?
A) RR = incidence exposed / incidence unexposed - cohort/RCT; OR = odds exposed / odds
unexposed - case-control, approximates RR when disease rare
B) Same thing always
C) OR only for cohort
D) RR only for case-control
CORRECT ANSWER: A) RR = incidence exposed / incidence unexposed - cohort/RCT;
OR = odds exposed / odds unexposed - case-control, approximates RR when disease
rare
RATIONALE: RR from cohort/RCT: risk ratio. OR from case-control: odds ratio, approximates RR when
disease rare (<10%). RR=1 no association, >1 risk, <1 protective.
QUESTION 8: PubH 6002: Which measure of central tendency best for skewed income
data?
A) Median - resistant to outliers, better for skewed
B) Mean always best
C) Mode always best
D) Range
CORRECT ANSWER: A) Median - resistant to outliers, better for skewed
RATIONALE: Mean affected by outliers/skewness, median resistant and better for skewed data like income.
Mode for categorical. For symmetric data mean=median=mode. Skewed right mean > median.
QUESTION 9: 68-95-99.7 rule for normal distribution means?
A) 68% within 1 SD, 95% within 2 SD, 99.7% within 3 SD of mean
B) 50% within 1 SD
PubH 6002 Biostatistics for Public Health | Practice Final | 2026/2027 UPDATE Page 4
