Biostatistics Practice Questions

Q: A screening test was positive in 50% of the diseased population and 10% of the healthy population. What is the specificity of the test?

0.9. ### **Explanation** To solve this question, we must apply the definitions of diagnostic test parameters to the percentages provided. 1. **Understanding the Data:** * **Sensitivity:** The probability that the test is positive in a diseased person. The question states the test is positive in **50%** of the diseased population. Thus, Sensitivity = 0.50. * **False Positive Rate (α):** The probability that the test is positive in a healthy person. The question states the test is positive in **10%** of the healthy population. Thus, False Positive Rate = 0.10. 2. **Calculating Specificity:** Specificity is the ability of a test to correctly identify those without the disease (True Negative Rate). It is mathematically related to the False Positive Rate: * **Specificity = 1 – False Positive Rate** * Specificity = 1 – 0.10 = **0.9 (or 90%)** --- ### **Analysis of Options:** * **Option B (0.9) is Correct:** As calculated above, specificity is the complement of the positive rate in healthy individuals (100% - 10% = 90%). * **Option A (0.5) is Incorrect:** This represents the **Sensitivity** of the test (positive results in the diseased population). * **Option C (0.83) is Incorrect:** This value does not correlate with the provided data; it is often a distractor representing a calculated Positive Predictive Value in specific prevalence scenarios. * **Option D (0.064) is Incorrect:** This is a mathematical distractor with no relevance to the basic definitions of sensitivity or specificity. --- ### **High-Yield Clinical Pearls for NEET-PG:** * **SNOUT:** **S**ensitivity helps rule **OUT** a disease (used in screening). * **SPIN:** **S**pecificity helps rule **IN** a disease (used in confirmation). * **False Positive Rate** is also known as **Type I Error (α)**. * **False Negative Rate** is also known as **Type II Error (β)**; Sensitivity is calculated as **(1 – β)**.

Q: If the prevalence of Candida glabrata infection is found to be 80% in a population of 100, what will be the range of prevalence at a 95% confidence level?

72% to 88%. ### Explanation **1. Understanding the Correct Answer (C):** The question asks for the **95% Confidence Interval (CI)** for a proportion. In biostatistics, the formula for the 95% Confidence Interval is: **$CI = p \pm 1.96 \times SE$** *(Where $p$ = prevalence/proportion and $SE$ = Standard Error)* * **Step 1:** Identify the variables. $p = 80\%$ (0.8), $q = 20\%$ (0.2), and $n = 100$. * **Step 2:** Calculate Standard Error (SE) for proportion: $SE = \sqrt{\frac{p \times q}{n}} = \sqrt{\frac{80 \times 20}{100}} = \sqrt{\frac{1600}{100}} = \sqrt{16} = 4$. * **Step 3:** Apply the 95% CI formula (using 2 as a rounded value for 1.96 for quick calculation): $80 \pm (2 \times 4) = 80 \pm 8$. * **Lower Limit:** $80 - 8 = 72\%$ * **Upper Limit:** $80 + 8 = 88\%$ Thus, we are 95% confident that the true population prevalence lies between **72% and 88%**. **2. Why Other Options are Incorrect:** * **Option A & D:** These ranges are too wide. They suggest a much larger Standard Error, which would only occur with a significantly smaller sample size (e.g., $n < 10$). * **Option B:** This range (65% to 95%) implies a Standard Error of 7.5, which does not mathematically align with the given sample size of 100. **3. High-Yield Clinical Pearls for NEET-PG:** * **Standard Error (SE):** Measures the precision of the sample estimate. As sample size ($n$) increases, SE decreases, and the Confidence Interval becomes narrower (more precise). * **Confidence Levels:** * 95% CI = $Mean \pm 1.96 \times SE$ (Commonly used in research) * 99% CI = $Mean \pm 2.58 \times SE$ * 68% CI = $Mean \pm 1 \times SE$ * **Prevalence vs. Incidence:** Prevalence (Total cases/Total population) is a cross-sectional measure, whereas Incidence (New cases/Population at risk) is a longitudinal measure.

Q: A study was conducted in a population of 2000 individuals. The mean hemoglobin level is 13.5 grams, and the distribution follows a normal distribution curve. What percentage of people will have a hemoglobin level more than 13.5 grams?

50%. ### Explanation **1. Why the Correct Answer (C) is Right:** The core concept here is the **Normal Distribution (Gaussian Distribution)**. In a perfectly normal distribution, the curve is symmetrical and bell-shaped. A fundamental property of this distribution is that the **Mean, Median, and Mode are all equal** and located exactly at the center of the curve. Since the Median represents the 50th percentile, exactly 50% of the observations lie below the mean and **50% lie above the mean**. In this question, the mean is 13.5 g/dL; therefore, regardless of the standard deviation or total population size (2000), 50% of the individuals will have a hemoglobin level higher than 13.5 g/dL. **2. Why the Incorrect Options are Wrong:** * **Option A (5%):** This value is associated with the "tails" of the distribution. In a normal distribution, approximately 5% of the population falls outside ±1.96 Standard Deviations (2.5% in each tail). * **Option B (25%) & D (75%):** These represent the First (Q1) and Third (Q3) Quartiles, respectively. While these are important markers in skewed distributions or box plots, they do not represent the division at the mean in a normal curve. **3. NEET-PG Clinical Pearls & High-Yield Facts:** * **Symmetry:** In a normal distribution, Skewness is **0** and Kurtosis is **3**. * **The 68-95-99.7 Rule (Empirical Rule):** * Mean ± 1 SD covers **68.2%** of values. * Mean ± 2 SD covers **95.4%** of values. * Mean ± 3 SD covers **99.7%** of values. * **Standard Normal Distribution:** A specific normal distribution where the Mean is **0** and the Standard Deviation is **1**. * **Z-score:** Indicates how many standard deviations a value is from the mean. At the mean (13.5 in this case), the Z-score is 0.

Q: Association can be measured by all of the following except?

Cronbach's alpha. ### Explanation The core concept tested here is the distinction between **measures of association** (how two variables relate) and **measures of reliability** (how consistently a tool measures a concept). **1. Why Cronbach’s Alpha is the Correct Answer:** Cronbach’s alpha is **not** a measure of association between two different variables. Instead, it is a measure of **internal consistency (reliability)**. It is used to determine how closely related a set of items are as a group (e.g., in a survey or a psychological test). A high alpha coefficient (usually >0.70) suggests that the items in a questionnaire are measuring the same underlying construct. **2. Analysis of Incorrect Options:** * **A. Correlation Coefficient (r):** This measures the strength and direction of a linear relationship between two quantitative variables (e.g., height and weight). It ranges from -1 to +1. * **C. P-value:** This is a measure of **statistical significance**. It indicates the probability that the observed association occurred by chance. While it doesn't measure the *strength* of association, it is a fundamental tool used to determine if an association exists. * **D. Odds Ratio (OR):** This is a classic measure of association used in **Case-Control studies**. It quantifies the odds of exposure in the diseased group compared to the non-diseased group. **High-Yield Clinical Pearls for NEET-PG:** * **Reliability vs. Validity:** Reliability is about *consistency* (Cronbach’s alpha); Validity is about *accuracy* (Sensitivity/Specificity). * **Relative Risk (RR):** Measure of association for **Cohort studies**. * **Attributable Risk (AR):** Measures the impact of an exposure on a population (clinical significance). * **Coefficient of Determination ($r^2$):** Tells you how much of the variation in one variable is explained by another.

Q: Population growth is considered less than adequate when the Net Reproduction Rate (NRR) is:

Less than 1. **Explanation:** The **Net Reproduction Rate (NRR)** is a demographic indicator that measures the average number of daughters a newborn girl will bear during her lifetime, assuming fixed age-specific fertility and mortality rates. It is the most relevant indicator for assessing population replacement. **1. Why Option A is Correct:** An **NRR < 1** indicates that each generation of mothers is failing to replace itself with at least one daughter. This leads to a decline in the population over time, signifying that population growth is **less than adequate** to maintain current levels. **2. Analysis of Incorrect Options:** * **Option B (NRR = 1):** This is known as **Replacement Level Fertility**. It means a mother is replaced by exactly one daughter who survives to reproductive age. This leads to a stable population (Zero Population Growth) in the long run. * **Option C (NRR > 1):** This indicates that the number of daughters born is greater than the number of mothers. This results in a **growing population**. * **Option D (Zero):** An NRR of zero would imply no female births or no females surviving to reproductive age, leading to eventual extinction, which is not a standard demographic growth classification. **Clinical Pearls & High-Yield Facts for NEET-PG:** * **NRR Goal:** The National Health Policy (NHP) of India aimed to achieve an **NRR of 1 by the year 2011** (a key demographic goal). * **NRR vs. GRR:** Unlike the Gross Reproduction Rate (GRR), the NRR **accounts for mortality** (the probability of a daughter surviving through her reproductive years). * **TFR Correlation:** To achieve an NRR of 1, the **Total Fertility Rate (TFR)** usually needs to be approximately **2.1** (Replacement level TFR). * **Formula:** $NRR = GRR \times \text{Probability of survival from birth to average age of childbearing}$.

Question 1

Randomization is done in clinical trials to:

Accepted Answer

Avoid selection bias within the study group

Answer

Avoid bias from the general population

Answer

Eliminate comparability

Answer

Ensure comparability

Question 2

A screening test was positive in 50% of the diseased population and 10% of the healthy population. What is the specificity of the test?

Accepted Answer

0.9

Answer

0.5

Answer

0.83

Answer

0.064

Question 3

Which of the following best represents a population with age variation?

Accepted Answer

Population pyramid

Answer

Life table

Answer

Correlation coefficient

Answer

Barchart

Question 4

If the prevalence of Candida glabrata infection is found to be 80% in a population of 100, what will be the range of prevalence at a 95% confidence level?

Accepted Answer

72% to 88%

Answer

4% to 100%

Answer

65% to 95%

Answer

70% to 100%

Question 5

A person is studying the relationship between deafness and mobile phone usage. Data on mobile phone usage was collected from a government source, and data on deafness was collected from district and private clinics. What type of study is this?

Accepted Answer

Ecological study

Answer

Experimental study

Answer

Etiological study

Answer

Cross-sectional study

Question 6

Which of the following biases can be eliminated by double blinding?

Accepted Answer

Interviewer's bias

Answer

Berksonian bias

Answer

Recall bias

Answer

Selection bias

Question 7

A study was conducted in a population of 2000 individuals. The mean hemoglobin level is 13.5 grams, and the distribution follows a normal distribution curve. What percentage of people will have a hemoglobin level more than 13.5 grams?

Accepted Answer

50%

Answer

5%

Answer

25%

Answer

75%

Question 8

Which of the following statements regarding sampling are true?

Accepted Answer

Snowball sampling is used for hidden populations.

Answer

More sample is required in systematic random sampling.

Answer

In stratified random sampling, the population is divided into strata.

Answer

Cluster sampling is less cost-effective.

Question 9

Association can be measured by all of the following except?

Accepted Answer

Cronbach's alpha

Answer

Correlation coefficient

Answer

P value

Answer

Odds ratio

Question 10

Population growth is considered less than adequate when the Net Reproduction Rate (NRR) is:

Accepted Answer

Less than 1

Answer

Equal to 1

Answer

Greater than 1

Answer

Zero

Biostatistics — MCQs

Biostatistics — MCQs

On this page

Practice by Chapter

Want unlimited practice?