WGU C784 - APPLIED HEALTHCARE STATISTICS PRE-
ASSESSMENT EXAM QUESTIONS AND DETAILED
ANSWERS - 188 Questions and Answers Already Graded A+
Premium Exam Tested And Verified
Subject Area Applied Healthcare Statistics
Description This exam assesses advanced competency in statistical methods applied to
healthcare, including probability distributions, hypothesis testing, regression
analysis, survival analysis, and data interpretation. It requires synthesis of
concepts and application to complex clinical and public health scenarios.
Expected Grade A+
Total Questions 188
Duration 3 hours
Learning Outcomes 1. Apply probability theory to healthcare decision-making
2. Interpret and evaluate statistical tests in medical literature
3. Perform and critique regression models for health outcomes
4. Analyze survival data using Kaplan-Meier and Cox regression
5. Design and interpret clinical trials and observational studies
Accreditation Conforms to U.S. university standards for upper-division statistics coursework,
including CAAHEP and AACSB guidelines for quantitative reasoning.
Page 1
,1. A clinical trial is designed to test a new anticoagulant against standard therapy for
preventing stroke in patients with atrial fibrillation. The primary endpoint is time to
first stroke. Which statistical method is most appropriate for comparing the two
treatment groups, adjusting for baseline covariates such as age and CHA2DS2-VASc
score?
A. Kaplan-Meier estimator with log-rank test
B. Cox proportional hazards regression
C. Logistic regression with odds ratio
D. Poisson regression for incidence rate ratio
Answer: B. Cox proportional hazards regression
Cox proportional hazards regression is the standard method for time-to-event data with
adjustment for covariates. The log-rank test (A) compares survival curves but does not
adjust for covariates. Logistic regression (C) is for binary outcomes at a fixed time,
ignoring time-to-event. Poisson regression (D) is for count data, not time-to-event with
censoring.
2. In a study of hospital readmission rates, researchers collect data on 5000 patients.
They use multiple linear regression to model length of stay (LOS) as a function of
age, comorbidity index, and number of prior admissions. The residual plot shows a
funnel shape (increasing variance with fitted values). Which assumption is violated,
and what is the most appropriate corrective action?
A. Normality; use robust standard errors
B. Homoscedasticity; apply a log transformation to LOS
C. Linearity; include an interaction term
D. Independence; use generalized estimating equations
Answer: B. Homoscedasticity; apply a log transformation to LOS
A funnel-shaped residual plot indicates heteroscedasticity (non-constant variance). Log
transformation often stabilizes variance for right-skewed outcomes like LOS. Robust
standard errors (A) address heteroscedasticity but do not fix the model; transformation
is more appropriate. Linearity (C) and independence (D) are not indicated by this
pattern.
Page 2
,3. A diagnostic test for a rare disease has a sensitivity of 95% and specificity of 90%.
The prevalence of the disease in the population is 1%. A patient tests positive. What
is the probability that the patient actually has the disease?
A. 0.0877
B. 0.0950
C. 0.5000
D. 0.9055
Answer: A. 0.0877
Using Bayes' theorem: P(Disease|Positive) = (0.95*0.01) / (0.95*0.01 + 0.10*0.99) =
0..1085 0.0877. Option B (0.095) is the positive predictive value if specificity
were 100%. Option C (0.5) might be a common misconception. Option D (0.9055) is the
negative predictive value.
4. A researcher wants to compare the mean systolic blood pressure among three
groups: control, low-dose drug, and high-dose drug. After collecting data, they
perform a one-way ANOVA and obtain a p-value of 0.04. However, the assumption
of homogeneity of variances is violated (Levene's test p = 0.01). Which of the
following is the most appropriate next step?
A. Proceed with ANOVA because it is robust to variance heterogeneity with large sample
sizes
B. Use Welch's ANOVA, which does not assume equal variances
C. Transform the data to achieve equal variances, then re-run ANOVA
D. Perform multiple t-tests with Bonferroni correction
Answer: B. Use Welch's ANOVA, which does not assume equal variances
Welch's ANOVA is designed for unequal variances and does not require the
homogeneity assumption. Option A is risky; robustness is not guaranteed with small or
moderate samples. Option C may be valid but is not the most direct; transformation
alters interpretation. Option D inflates Type I error and does not address variance
issue.
Page 3
, 5. In a cohort study examining the association between occupational exposure to
asbestos and lung cancer mortality, the hazard ratio (HR) from a Cox model is 2.5
(95% CI: 1.8-3.5). Which interpretation is correct?
A. The exposed group has 2.5 times the risk of dying from lung cancer at any given time
compared to the unexposed group
B. The exposed group has 2.5 times the odds of lung cancer death compared to the
unexposed group
C. The exposed group has 2.5 times the incidence rate of lung cancer death over the study
period
D. The exposed group has a 250% higher risk of lung cancer death at the end of follow-up
Answer: A. The exposed group has 2.5 times the risk of dying from lung cancer at
any given time compared to the unexposed group
The hazard ratio represents the instantaneous risk of the event occurring in the exposed
group relative to the unexposed at any point in time. Option B refers to odds ratio, not
hazard. Option C describes rate ratio but not instantaneous. Option D implies a
cumulative risk difference, which is not directly given by HR.
6. A medical researcher wants to test whether a new biomarker predicts
progression-free survival in cancer patients, after adjusting for age, tumor stage, and
performance status. The outcome is time to progression or death, and there is
substantial censoring. Which modeling approach is most appropriate?
A. Multiple linear regression on log-transformed survival time
B. Cox proportional hazards regression
C. Binary logistic regression for progression status at 1 year
D. Poisson regression for number of progressions
Answer: B. Cox proportional hazards regression
Cox regression is specifically designed for time-to-event data with censoring and allows
adjustment for covariates. Linear regression (A) cannot handle censoring properly.
Logistic regression (C) ignores time and censoring. Poisson regression (D) is for count
data, not time-to-event.
Page 4
ASSESSMENT EXAM QUESTIONS AND DETAILED
ANSWERS - 188 Questions and Answers Already Graded A+
Premium Exam Tested And Verified
Subject Area Applied Healthcare Statistics
Description This exam assesses advanced competency in statistical methods applied to
healthcare, including probability distributions, hypothesis testing, regression
analysis, survival analysis, and data interpretation. It requires synthesis of
concepts and application to complex clinical and public health scenarios.
Expected Grade A+
Total Questions 188
Duration 3 hours
Learning Outcomes 1. Apply probability theory to healthcare decision-making
2. Interpret and evaluate statistical tests in medical literature
3. Perform and critique regression models for health outcomes
4. Analyze survival data using Kaplan-Meier and Cox regression
5. Design and interpret clinical trials and observational studies
Accreditation Conforms to U.S. university standards for upper-division statistics coursework,
including CAAHEP and AACSB guidelines for quantitative reasoning.
Page 1
,1. A clinical trial is designed to test a new anticoagulant against standard therapy for
preventing stroke in patients with atrial fibrillation. The primary endpoint is time to
first stroke. Which statistical method is most appropriate for comparing the two
treatment groups, adjusting for baseline covariates such as age and CHA2DS2-VASc
score?
A. Kaplan-Meier estimator with log-rank test
B. Cox proportional hazards regression
C. Logistic regression with odds ratio
D. Poisson regression for incidence rate ratio
Answer: B. Cox proportional hazards regression
Cox proportional hazards regression is the standard method for time-to-event data with
adjustment for covariates. The log-rank test (A) compares survival curves but does not
adjust for covariates. Logistic regression (C) is for binary outcomes at a fixed time,
ignoring time-to-event. Poisson regression (D) is for count data, not time-to-event with
censoring.
2. In a study of hospital readmission rates, researchers collect data on 5000 patients.
They use multiple linear regression to model length of stay (LOS) as a function of
age, comorbidity index, and number of prior admissions. The residual plot shows a
funnel shape (increasing variance with fitted values). Which assumption is violated,
and what is the most appropriate corrective action?
A. Normality; use robust standard errors
B. Homoscedasticity; apply a log transformation to LOS
C. Linearity; include an interaction term
D. Independence; use generalized estimating equations
Answer: B. Homoscedasticity; apply a log transformation to LOS
A funnel-shaped residual plot indicates heteroscedasticity (non-constant variance). Log
transformation often stabilizes variance for right-skewed outcomes like LOS. Robust
standard errors (A) address heteroscedasticity but do not fix the model; transformation
is more appropriate. Linearity (C) and independence (D) are not indicated by this
pattern.
Page 2
,3. A diagnostic test for a rare disease has a sensitivity of 95% and specificity of 90%.
The prevalence of the disease in the population is 1%. A patient tests positive. What
is the probability that the patient actually has the disease?
A. 0.0877
B. 0.0950
C. 0.5000
D. 0.9055
Answer: A. 0.0877
Using Bayes' theorem: P(Disease|Positive) = (0.95*0.01) / (0.95*0.01 + 0.10*0.99) =
0..1085 0.0877. Option B (0.095) is the positive predictive value if specificity
were 100%. Option C (0.5) might be a common misconception. Option D (0.9055) is the
negative predictive value.
4. A researcher wants to compare the mean systolic blood pressure among three
groups: control, low-dose drug, and high-dose drug. After collecting data, they
perform a one-way ANOVA and obtain a p-value of 0.04. However, the assumption
of homogeneity of variances is violated (Levene's test p = 0.01). Which of the
following is the most appropriate next step?
A. Proceed with ANOVA because it is robust to variance heterogeneity with large sample
sizes
B. Use Welch's ANOVA, which does not assume equal variances
C. Transform the data to achieve equal variances, then re-run ANOVA
D. Perform multiple t-tests with Bonferroni correction
Answer: B. Use Welch's ANOVA, which does not assume equal variances
Welch's ANOVA is designed for unequal variances and does not require the
homogeneity assumption. Option A is risky; robustness is not guaranteed with small or
moderate samples. Option C may be valid but is not the most direct; transformation
alters interpretation. Option D inflates Type I error and does not address variance
issue.
Page 3
, 5. In a cohort study examining the association between occupational exposure to
asbestos and lung cancer mortality, the hazard ratio (HR) from a Cox model is 2.5
(95% CI: 1.8-3.5). Which interpretation is correct?
A. The exposed group has 2.5 times the risk of dying from lung cancer at any given time
compared to the unexposed group
B. The exposed group has 2.5 times the odds of lung cancer death compared to the
unexposed group
C. The exposed group has 2.5 times the incidence rate of lung cancer death over the study
period
D. The exposed group has a 250% higher risk of lung cancer death at the end of follow-up
Answer: A. The exposed group has 2.5 times the risk of dying from lung cancer at
any given time compared to the unexposed group
The hazard ratio represents the instantaneous risk of the event occurring in the exposed
group relative to the unexposed at any point in time. Option B refers to odds ratio, not
hazard. Option C describes rate ratio but not instantaneous. Option D implies a
cumulative risk difference, which is not directly given by HR.
6. A medical researcher wants to test whether a new biomarker predicts
progression-free survival in cancer patients, after adjusting for age, tumor stage, and
performance status. The outcome is time to progression or death, and there is
substantial censoring. Which modeling approach is most appropriate?
A. Multiple linear regression on log-transformed survival time
B. Cox proportional hazards regression
C. Binary logistic regression for progression status at 1 year
D. Poisson regression for number of progressions
Answer: B. Cox proportional hazards regression
Cox regression is specifically designed for time-to-event data with censoring and allows
adjustment for covariates. Linear regression (A) cannot handle censoring properly.
Logistic regression (C) ignores time and censoring. Poisson regression (D) is for count
data, not time-to-event.
Page 4
