Biostatistics Practice Questions

Q: The regression between height and age follows the equation y = a + bx. What type of curve does this represent?

Straight line. ### Explanation **1. Why "Straight Line" is Correct:** The equation **y = a + bx** is the standard mathematical formula for a **Simple Linear Regression**. * **y** is the dependent variable (e.g., Height). * **x** is the independent variable (e.g., Age). * **a** is the intercept (the value of y when x is zero). * **b** is the regression coefficient (the slope of the line). In biostatistics, linear regression is used to predict the value of one continuous variable based on another. Because the power of the variable 'x' is 1 (first-degree equation), the relationship plotted on a Cartesian plane results in a **straight line**. **2. Why Other Options are Incorrect:** * **A. Hyperbola:** This represents an inverse relationship (y = 1/x) where one variable increases as the other decreases at a non-linear rate. * **B. Sigmoid:** This is an S-shaped curve common in **Logistic Regression**, used when the dependent variable is binary (e.g., Dead/Alive or Diseased/Healthy). * **C. Parabola:** This represents a quadratic relationship ($y = ax^2 + bx + c$), where the direction of the curve changes once. **3. Clinical Pearls & High-Yield Facts for NEET-PG:** * **Correlation vs. Regression:** Correlation ($r$) measures the *strength and direction* of a relationship, while Regression ($b$) allows for the *prediction* of one variable from another. * **Coefficient of Determination ($r^2$):** This indicates the proportion of variance in the dependent variable that is predictable from the independent variable. * **Range of $r$:** Correlation coefficient ranges from -1 to +1, whereas the regression coefficient ($b$) can range from $-\infty$ to $+\infty$. * **Scatter Diagram:** This is the best visual method to initially assess the relationship between two quantitative variables before calculating regression.

Q: Equal interval reduces which bias?

Selection bias. ### Explanation **1. Why Selection Bias is the Correct Answer** Selection bias occurs when the study population is not representative of the target population due to the way subjects are chosen. **Equal interval** refers to **Systematic Random Sampling**, where subjects are selected at fixed, predetermined intervals (e.g., every 5th person on a list). By using a fixed mathematical interval starting from a random point, the researcher eliminates subjective judgment and ensures that every segment of the sampling frame is represented. This objective approach minimizes the risk of the investigator "cherry-picking" specific participants, thereby significantly reducing **Selection Bias**. **2. Analysis of Incorrect Options** * **B. Berksonian Bias:** This is a specific type of selection bias (also called admission rate bias) that occurs in hospital-based case-control studies because hospitalized patients have different exposure rates than the general population. Equal intervals in a general sampling frame do not address the inherent bias of using a hospital population. * **C. Interviewer Bias:** This is a type of **Information (Observation) Bias**. it occurs during the data collection phase when the interviewer’s subconscious hand-gestures or leading questions influence the respondent's answer. It is minimized by "blinding" or standardized questionnaires, not by the sampling interval. **3. High-Yield Clinical Pearls for NEET-PG** * **Systematic Sampling:** The interval is calculated as $K = N/n$ (where $N$ = Total Population and $n$ = Sample Size). * **Gold Standard:** While systematic sampling reduces bias, **Simple Random Sampling** (using random number tables) is often considered the best for eliminating selection bias. * **Recall Bias:** Most common in Case-Control studies; minimized by using objective records. * **Confounding:** Minimized at the design stage by **Randomization, Matching, and Restriction**.

Q: True positivity is indicated by which of the following measures?

Sensitivity. **Explanation:** **Sensitivity** is defined as the ability of a test to correctly identify those who have the disease. It represents the **True Positive Rate**. Mathematically, it is calculated as: `Sensitivity = [True Positives (TP) / (True Positives + False Negatives)] × 100`. In clinical practice, a highly sensitive test is used for screening because it ensures that very few cases are missed (low false negatives). **Analysis of Incorrect Options:** * **B. Specificity:** This measures the **True Negativity**. It is the ability of a test to correctly identify those without the disease. It is calculated as `[True Negatives (TN) / (True Negatives + False Positives)]`. * **C. Predictive Value:** This refers to the probability that a person with a positive test result actually has the disease (Positive Predictive Value) or a person with a negative result is truly healthy (Negative Predictive Value). It depends heavily on the **prevalence** of the disease in the population. * **D. Validity:** This is a broader term indicating the accuracy of a test—the degree to which a test measures what it intends to measure. It encompasses both sensitivity and specificity. **High-Yield Pearls for NEET-PG:** * **SNOUT:** **S**ensitivity rules **OUT** disease (when the result is negative). * **SPIN:** **S**pecificity rules **IN** disease (when the result is positive). * Sensitivity and Specificity are **inherent properties** of a test and do not change with disease prevalence. * In contrast, **Predictive Values** are inversely or directly proportional to prevalence (PPV increases with prevalence; NPV decreases).

Q: If the cut-off point for blood glucose is lowered to detect diabetes (e.g., less than 120 mg/dL), what is the effect on the test's performance characteristics?

Sensitivity of the test is increased. ### Explanation **1. Why Option A is Correct:** In screening, sensitivity and specificity have an **inverse relationship**. When you lower the cut-off point for a continuous variable (like blood glucose), you are "widening the net." By lowering the threshold to <120 mg/dL, you will capture more people who actually have the disease (True Positives), but you will also capture more healthy people as "diseased." * **Concept:** Lowering the cut-off increases the number of **True Positives**, thereby **increasing Sensitivity**. **2. Why Other Options are Incorrect:** * **Option B:** Specificity is the ability of a test to correctly identify those without the disease. By lowering the cut-off, you increase the number of **False Positives** (healthy people labeled as diabetic). Since Specificity = TN / (TN + FP), an increase in False Positives leads to a **decrease in Specificity**. * **Option C:** "Predictive accuracy" usually refers to Predictive Values (PPV/NPV). When the cut-off is lowered, the **Positive Predictive Value (PPV) decreases** because the number of False Positives rises significantly. However, the question specifically asks for the primary effect on performance characteristics; the most direct and certain impact is on Sensitivity. **3. Clinical Pearls & High-Yield Facts for NEET-PG:** * **Mnemonic:** **L**ower the cut-off = **L**ess Specific, but more Sensitive. * **Screening vs. Diagnostic:** Screening tests require high **Sensitivity** (to not miss cases), while diagnostic tests require high **Specificity** (to confirm the disease). * **The "Trade-off":** On a ROC (Receiver Operating Characteristic) curve, moving the cut-off point to the left/down increases sensitivity at the cost of specificity. * **Prevalence:** Remember that while Sensitivity and Specificity are inherent properties of the test, **Predictive Values** (PPV/NPV) change with the prevalence of the disease in the population.

Q: Which of the following graphical methods is best for studying the decline in the percentage of alcohol usage in men and women over several years?

Line diagram. ### Explanation **1. Why the Correct Answer is Right:** The **Line Diagram** (or Line Graph) is the most suitable method for representing **trends over time**. In biostatistics, when we need to observe the progression, decline, or fluctuation of a variable (like alcohol usage percentage) across several years, a line diagram effectively connects data points to show the direction of change. It is particularly useful for comparing two or more groups (men vs. women) on the same axes, allowing for a clear visual comparison of their respective trends. **2. Why Other Options are Incorrect:** * **Pie Chart:** This is used to show the **proportional distribution** of a single variable at a specific point in time (e.g., the share of different types of substances used). It cannot represent changes over a time series. * **Histogram:** This is used for **continuous quantitative data** to show frequency distribution within a single group. It consists of adjacent rectangles where the area represents the frequency. It is not designed to show trends over years. * **Frequency Polygon:** This is a derivative of the histogram, created by joining the midpoints of the tops of the histogram bars. While it shows distribution, it is used for frequency data, not for tracking a percentage trend over a chronological period. **3. Clinical Pearls & High-Yield Facts for NEET-PG:** * **Trend over time:** Always choose **Line Diagram**. * **Correlation between two variables:** Always choose **Scatter Diagram**. * **Comparison of discrete/qualitative data:** Use a **Bar Chart**. * **Frequency distribution of continuous data:** Use a **Histogram**. * **To find Median graphically:** Use an **Ogive** (Cumulative frequency curve). * **Pictogram:** Uses images to represent data; easiest for a layperson to understand but least accurate.

Q: All of the following features are true about a cross-sectional study EXCEPT?

A cause and effect relationship can be established. ### Explanation **Why Option D is the Correct Answer (The "Except" Statement):** A cross-sectional study is often referred to as a **"Snapshot Study"** because it measures both the exposure and the outcome simultaneously at a single point in time. Because both are measured together, it is impossible to determine whether the exposure preceded the outcome. This lack of **temporality** (the first criterion of Bradford Hill’s criteria for causation) means that a cross-sectional study can show an **association** but cannot establish a definitive **cause-and-effect relationship**. **Analysis of Incorrect Options:** * **Option A:** This is a defining feature. It examines the prevalence of a condition in a population at one specific point in time (like a still photograph). * **Option B:** Unlike cohort studies, there is no forward-looking component. Once the data is collected from the participants, the study is complete; hence, follow-up is not required. * **Option C:** These studies are ideal for chronic conditions (e.g., Diabetes, Hypertension) because they measure **prevalence**. They are less useful for acute or rare diseases, as the "snapshot" might miss short-lived cases. **High-Yield Clinical Pearls for NEET-PG:** * **Prevalence Study:** Cross-sectional study is the best design to calculate the prevalence of a disease. * **Hypothesis Generation:** It is used to generate hypotheses, whereas analytical studies (Case-control/Cohort) are used to test them. * **Sequence of Strength:** Cross-sectional < Case-Control < Cohort < Randomized Controlled Trial (RCT). * **Ecological Fallacy:** Often associated with ecological studies, but remember that cross-sectional studies focus on individuals, not populations.

Q: If the mean, median, and mode are 10, 18, and 26 respectively, what kind of distribution is it?

Negatively skewed. ### Explanation **1. Why the Correct Answer is Right:** In biostatistics, the relationship between the **Mean, Median, and Mode** determines the shape of a frequency distribution curve. * In this question: **Mean (10) < Median (18) < Mode (26).** * When the mean is pulled toward the lower values (the left side), it indicates a **Negatively Skewed Distribution** (also known as "Left-skewed"). * The "tail" of the graph points toward the smaller numbers (negative side of the X-axis). This happens when there are a few extremely low values that drag the mean down, while the majority of data points are clustered at the higher end. **2. Why the Incorrect Options are Wrong:** * **Symmetric / Normal Distribution (Options A & B):** In a perfectly symmetrical or Normal (Gaussian) distribution, the **Mean = Median = Mode**. Since 10 ≠ 18 ≠ 26, the distribution is asymmetrical. * **Positively Skewed (Option C):** In a positively skewed distribution (Right-skewed), the relationship is reversed: **Mean > Median > Mode**. The tail points toward the higher values (positive side), usually due to a few extremely high outliers. **3. NEET-PG Clinical Pearls & High-Yield Facts:** * **The "Alphabetical Rule":** To remember the order in a **Positively Skewed** distribution, follow the alphabet: **M**ean > **M**edian > **M**ode (alphabetical order of the second letters: e > i > o). * **Median's Position:** In any skewed distribution (positive or negative), the **Median always stays in the middle** between the Mean and the Mode. * **Best Measure of Central Tendency:** * For **Normal Distribution**: Mean is the best measure. * For **Skewed Distribution**: Median is the best measure (as it is not affected by extreme outliers). * **Formula (Karl Pearson’s):** $Mode = (3 \times Median) - (2 \times Mean)$. This is often used to calculate a missing value in NEET-PG numericals.

Q: At the end of the year 1990, the population of a primary health centre was 30,000 and there were 120 cases of pulmonary tuberculosis. At the end of the year 1991, the population was 30,600, 30 new cases were detected, and 2 cases had died. Based on this data, all of the following rates can be calculated EXCEPT?

Proportional mortality rate. To solve this question, we must understand the specific data requirements for various epidemiological indicators. ### **Why Proportional Mortality Rate (PMR) is the Correct Answer** **Proportional Mortality Rate** is defined as the number of deaths due to a specific cause (e.g., TB) divided by the **total number of deaths** from all causes in that population during the same period. * **The Missing Link:** While we know 2 people died of TB, the question does not provide the **total deaths from all causes** in the PHC. Without this denominator, PMR cannot be calculated. ### **Analysis of Incorrect Options** * **A. Incidence Rate:** This is the number of **new cases** (30) occurring in a population at risk during a specific period. Since we have the new cases and the mid-year population (approx. 30,300), this can be calculated. * **B. Prevalence Rate:** This includes **all cases** (old + new) existing at a point in time. At the end of 1991, the prevalence would be (120 old + 30 new - 2 deaths) divided by the total population. Thus, it is calculable. * **C. Case Fatality Rate (CFR):** This measures the killing power of a disease. It is calculated as (Deaths due to TB / Total cases of TB) × 100. Here, (2 / 150) × 100 provides the CFR. ### **High-Yield Clinical Pearls for NEET-PG** 1. **Incidence vs. Prevalence:** Incidence is a measure of **rate** (new cases/waterfall), while Prevalence is a measure of **proportion** (total cases/pool). 2. **PMR vs. CFR:** PMR indicates the *burden* of a specific disease relative to all deaths in a community, whereas CFR indicates the *severity* of the disease among those afflicted. 3. **Denominator Check:** Always look for "Total Deaths" for PMR and "Total Population" for Crude Death Rate. If neither is present, mortality rates (except CFR) usually cannot be calculated.

Question 1

The regression between height and age follows the equation y = a + bx. What type of curve does this represent?

Accepted Answer

Straight line

Answer

Hyperbola

Answer

Sigmoid

Answer

Parabola

Question 2

Equal interval reduces which bias?

Accepted Answer

Selection bias

Answer

Berksonian bias

Answer

Interviewer bias

Answer

None

Question 3

True positivity is indicated by which of the following measures?

Accepted Answer

Sensitivity

Answer

Specificity

Answer

Predictive value

Answer

Validity

Question 4

Which of the following statements about Simple Random Sampling is true?

Accepted Answer

Every element in the population has an equal probability of being included.

Answer

Sampling is based on similar characteristics.

Answer

Suitable for large heterogeneous population.

Answer

A complete list of items within the sampling frame is not required.

Question 5

If the cut-off point for blood glucose is lowered to detect diabetes (e.g., less than 120 mg/dL), what is the effect on the test's performance characteristics?

Accepted Answer

Sensitivity of the test is increased

Answer

Specificity of the test is increased

Answer

Predictive accuracy of the test is decreased

Answer

None of the above

Question 6

Which of the following graphical methods is best for studying the decline in the percentage of alcohol usage in men and women over several years?

Accepted Answer

Line diagram

Answer

Pie chart

Answer

Histogram

Answer

Frequency polygon

Question 7

All of the following features are true about a cross-sectional study EXCEPT?

Accepted Answer

A cause and effect relationship can be established

Answer

All cases are seen at one point in time

Answer

Follow-up is not a necessary feature

Answer

More useful for chronic diseases

Question 8

If the mean, median, and mode are 10, 18, and 26 respectively, what kind of distribution is it?

Accepted Answer

Negatively skewed

Answer

Symmetric

Answer

Normal

Answer

Positively skewed

Question 9

At the end of the year 1990, the population of a primary health centre was 30,000 and there were 120 cases of pulmonary tuberculosis. At the end of the year 1991, the population was 30,600, 30 new cases were detected, and 2 cases had died. Based on this data, all of the following rates can be calculated EXCEPT?

Accepted Answer

Proportional mortality rate

Answer

Incidence rate

Answer

Prevalence rate

Answer

Case fatality rate

Question 10

If the specificity of a diagnostic test increases, what happens to the number of false negatives?

Accepted Answer

False negatives increase

Answer

False negatives decrease

Answer

True negatives decrease

Answer

False positives increase

Biostatistics — MCQs

Biostatistics — MCQs

On this page

Practice by Chapter

Want unlimited practice?