Biostatistics Practice Questions

Q: Which type of variable best describes the number of family members?

Discrete variable. **Explanation:** In Biostatistics, variables are classified based on the nature of the data they represent. The **number of family members** is a classic example of a **Discrete Variable**. **1. Why Discrete Variable is correct:** A discrete variable is a type of quantitative (numerical) variable that can only take on specific, whole-number values. These values are obtained by **counting**. Since you cannot have 4.5 or 5.2 family members—only 4, 5, or 6—the data exists in distinct, separate units with no possible values in between. **2. Why other options are incorrect:** * **Continuous Variable:** These are numerical variables that can take any value within a range, including decimals and fractions. They are typically obtained by **measuring**. Examples include height (165.5 cm), weight (70.2 kg), or blood pressure. * **Qualitative / Categorical Variable:** These describe a quality or attribute rather than a numerical quantity. They are expressed in words or categories. Examples include gender (Male/Female), blood group (A, B, AB, O), or socioeconomic status. While "number of family members" can be grouped into categories (e.g., Small vs. Large), the number itself is inherently quantitative. **Clinical Pearls for NEET-PG:** * **Memory Aid:** **D**iscrete = **D**isconnected (whole numbers); **C**ontinuous = **C**onnected (decimals possible). * **Scales of Measurement:** Discrete and Continuous variables fall under **Interval** or **Ratio** scales. * **High-Yield Example:** Number of hospital beds or number of cases of a disease are **Discrete**; Hemoglobin levels or Serum Creatinine are **Continuous**.

Q: In a cohort study of 7000 smokers over ten years, 70 developed lung cancer. In a concurrent evaluation of 7000 non-smokers in the same catchment area, 7 developed lung cancer. What is the Relative Risk (RR) for developing lung cancer?

10. ### Explanation **1. Why Option B (10) is Correct:** Relative Risk (RR) is the ratio of the incidence of a disease among an exposed group to the incidence among a non-exposed group. It is the primary measure of association in **Cohort Studies**. * **Incidence in Exposed (Smokers):** $I_e = \frac{\text{New cases}}{\text{Total exposed}} = \frac{70}{7000} = 0.01$ (or 10 per 1000) * **Incidence in Non-exposed (Non-smokers):** $I_o = \frac{\text{New cases}}{\text{Total non-exposed}} = \frac{7}{7000} = 0.001$ (or 1 per 1000) * **Formula for Relative Risk (RR):** $\frac{I_e}{I_o} = \frac{0.01}{0.001} = \mathbf{10}$ This means smokers are 10 times more likely to develop lung cancer compared to non-smokers. **2. Why Other Options are Incorrect:** * **Option A (1):** An RR of 1 indicates "Null Hypothesis" (no association between exposure and disease). * **Option C (100):** This would imply a much higher strength of association, likely due to a calculation error in decimal placement. * **Option D (0.1):** An RR < 1 indicates a "Protective Effect" (the exposure prevents the disease), which is clinically incorrect for smoking and cancer. **3. High-Yield Clinical Pearls for NEET-PG:** * **Relative Risk (RR):** Direct measure of the **strength of association**. It is calculated only in prospective studies (Cohort). * **Odds Ratio (OR):** Used in Case-Control studies as an estimate of RR. * **Attributable Risk (AR):** $(I_e - I_o) / I_e \times 100$. It indicates the amount of disease that can be prevented if the exposure is eliminated. * **Population Attributable Risk (PAR):** Useful for public health administrators to prioritize interventions in the community.

Q: The "Crude death rate" is defined as the number of deaths (from all causes) per 1000 estimated what?

Mid-year population. **Explanation:** **1. Why "Mid-year population" is correct:** The Crude Death Rate (CDR) is a fundamental measure of mortality in a population. It is calculated as the number of deaths occurring during a calendar year per 1000 of the **mid-year population**. The mid-year population (estimated as of July 1st) is used as the denominator because it represents the "average" population at risk of dying throughout that year, accounting for births, deaths, and migrations that occur during the 12-month period. **2. Why other options are incorrect:** * **Total population:** While CDR relates to the population, "Total population" is vague. In demography, the population size fluctuates daily; therefore, the specific mid-year estimate is the standardized denominator used for annual rates. * **Total births / Live births:** These are used as denominators for mortality indicators specifically related to early life, such as the **Infant Mortality Rate (IMR)** or **Maternal Mortality Ratio (MMR)**, rather than the general death rate of the entire community. **3. NEET-PG High-Yield Pearls:** * **Formula:** $CDR = \frac{\text{Number of deaths during the year}}{\text{Mid-year population}} \times 1000$. * **Limitation:** The CDR is "crude" because it does not account for the age and sex composition of the population. A population with many elderly individuals will have a higher CDR than a younger population, even if health conditions are better. * **Comparison:** To compare mortality between two different populations (e.g., Kerala vs. UP), **Age-Standardized Death Rates** are the preferred indicator to eliminate the bias of age distribution. * **Current Trend:** According to recent SRS (Sample Registration System) data, the CDR for India is approximately **6.0 per 1000** mid-year population.

Q: Type 1 Sampling error is classified as?

Alpha error. In biostatistics, hypothesis testing involves making a decision about a population based on sample data. Errors occur when this decision does not reflect the true state of the population. ### **Explanation of the Correct Answer** **A. Alpha (α) Error (Type I Error):** This occurs when a researcher **rejects a null hypothesis that is actually true**. In clinical terms, it is a "False Positive" result—concluding that a treatment works or a difference exists when, in reality, it does not. The probability of committing a Type I error is denoted by the significance level (α), commonly set at 0.05 (5%). ### **Explanation of Incorrect Options** * **B. Beta (β) Error (Type II Error):** This occurs when a researcher **fails to reject a null hypothesis that is actually false**. It is a "False Negative" result—concluding there is no difference when one actually exists. * **C & D. Gamma and Delta Errors:** These are not standard terms used to classify primary sampling errors in classical hypothesis testing. While "Gamma" is used in specific correlation coefficients and "Delta" often represents effect size, they do not describe Type I or II errors. ### **NEET-PG High-Yield Pearls** * **Confidence Level:** Calculated as **(1 – α)**. It represents the probability of correctly accepting a true null hypothesis. * **Power of a Study:** Calculated as **(1 – β)**. It is the ability of a study to detect a difference if one truly exists. To increase power, one should increase the sample size. * **P-value:** The probability of committing a Type I error. If p < 0.05, the result is statistically significant. * **Memory Aid:** * **Type I (α):** **I**nnocent person goes to jail (False Positive). * **Type II (β):** **B**ad person goes free (False Negative).

Q: A village is divided into five relevant subgroups for the purpose of a survey. Individuals from each subgroup are then selected randomly. What is this type of sampling called?

Stratified sampling. ### Explanation **1. Why Stratified Sampling is Correct:** In **Stratified Random Sampling**, the heterogeneous population is first divided into non-overlapping, homogeneous subgroups called **"strata"** based on specific characteristics (e.g., age, gender, socio-economic status, or "relevant subgroups" as mentioned in the question). A **simple random sample** is then drawn from *each* of these strata. This ensures that every subgroup is adequately represented, reducing sampling error compared to simple random sampling. **2. Why Other Options are Incorrect:** * **Simple Random Sampling:** Every individual in the entire population has an equal chance of being selected. There is no prior division into subgroups. * **Cluster Sampling:** The population is divided into groups (clusters), usually based on geographical areas (e.g., villages, wards). Unlike stratified sampling, you randomly select a few *entire clusters* and survey everyone within them, rather than selecting individuals from every group. * **Systematic Sampling:** This involves selecting every $k^{th}$ individual (sampling interval) from a list, starting from a random point (e.g., every 5th person entering an OPD). **3. High-Yield Clinical Pearls for NEET-PG:** * **Stratified vs. Cluster:** In Stratified sampling, the groups are **homogeneous within** (similar people) but **heterogeneous between** (strata differ from each other). In Cluster sampling, groups are **heterogeneous within** but **homogeneous between** (each cluster is a mini-reflection of the population). * **Multistage Sampling:** This is the most common method used in large-scale national health surveys (like NFHS), involving a combination of sampling techniques. * **Precision:** Stratified sampling is generally more precise than simple random sampling because it accounts for variability between subgroups.

Q: What is true about the standard normal distribution?

The total area under the curve is equal to 1.. ### Explanation **1. Why Option A is Correct:** The Standard Normal Distribution (Z-distribution) is a specific type of probability density function. In statistics, the **total area under any probability curve must equal 1 (or 100%)**, representing the sum of all possible outcomes. This property is fundamental for calculating Z-scores and p-values, as the area under specific segments of the curve represents the probability of an observation falling within that range. **2. Why the Other Options are Incorrect:** * **Option B:** In a *Standard* Normal Distribution, the **Mean is always 0** and the Standard Deviation is 1. If the mean were 1, it would simply be a "Normal Distribution," not the "Standard" version. * **Option C:** The Normal Distribution is perfectly symmetrical. Therefore, the **Mean = Median = Mode**. The relationship "Mean > Median > Mode" describes a **Positively Skewed** distribution. * **Option D:** A distribution with a tail towards the right is **Positively Skewed**. The Standard Normal Distribution is bell-shaped and symmetrical with no skew; both tails extend infinitely but are identical in shape. **3. High-Yield Clinical Pearls for NEET-PG:** * **Z-score formula:** $Z = (x - \mu) / \sigma$. It tells you how many standard deviations a value is from the mean. * **Empirical Rule (68-95-99.7 Rule):** * Mean ± 1 SD covers **68.2%** of the area. * Mean ± 2 SD covers **95.4%** of the area. * Mean ± 3 SD covers **99.7%** of the area. * **Point of Inflection:** In a normal curve, this occurs at Mean ± 1 SD (where the curve changes from convex to concave). * **Standard Error:** As sample size increases, the standard error decreases, making the distribution narrower.

Q: Which of the following is true about a cohort study?

Incidence can be calculated. **Explanation:** A **Cohort Study** is an observational, analytical study design where a group of individuals (the cohort) is defined based on the presence or absence of exposure to a particular factor and followed forward in time to observe the development of an outcome. **Why Option A is correct:** In a cohort study, we start with a population that is **at risk** but free of the disease. Since we follow these individuals over a period to see how many new cases develop, we can directly measure the **Incidence** (number of new cases in a population at risk). This allows for the calculation of **Relative Risk (RR)** and **Attributable Risk (AR)**. **Why other options are incorrect:** * **Option B:** Cohort studies proceed from **Cause to Effect** (Prospective). It is the Case-Control study that proceeds from effect (disease) to cause (exposure). * **Option C:** Cohort studies are generally **expensive** because they require large sample sizes and long-term follow-up of healthy individuals. * **Option D:** They require a **longer time** than case-control studies because the investigator must wait for the disease to manifest after exposure. **High-Yield Clinical Pearls for NEET-PG:** * **Best for:** Rare exposures (not rare diseases). * **Key Measure:** Relative Risk (RR). If RR > 1, there is a positive association. * **Bias:** The most common bias in cohort studies is **Attrition Bias** (loss to follow-up). * **Nested Case-Control Study:** A case-control study conducted within a large cohort study; it is more economical and reduces selection bias.

Q: Interpret the statistical graph shown below:

Positive correlation. ***Positive correlation*** - In a **positive correlation**, as one variable increases, the other variable also increases, creating an **upward trending pattern** on the scatter plot. - The data points form a **linear pattern** sloping from bottom-left to top-right, indicating a **direct relationship** between the variables. *Negative correlation* - Shows a **downward trending pattern** where one variable increases while the other decreases. - Data points slope from **top-left to bottom-right**, indicating an **inverse relationship** between variables. *Absent correlation* - Data points are **randomly scattered** with no discernible pattern or trend. - The **correlation coefficient (r)** approaches **zero**, indicating no linear relationship between variables. *Spurious correlation* - Represents a **false association** between two variables that appear correlated but lack a true causal relationship. - Often occurs due to **confounding variables** or **coincidental patterns** in the data.

Question 1

Which type of variable best describes the number of family members?

Accepted Answer

Discrete variable

Answer

Qualitative variable

Answer

Continuous variable

Answer

Categorical variable

Question 2

Which scale is used for classifying data that lacks a particular structure and is presented without inherent order?

Accepted Answer

Nominal

Answer

Ordinal

Answer

Interval

Answer

Ratio

Question 3

In a cohort study of 7000 smokers over ten years, 70 developed lung cancer. In a concurrent evaluation of 7000 non-smokers in the same catchment area, 7 developed lung cancer. What is the Relative Risk (RR) for developing lung cancer?

Accepted Answer

10

Answer

1

Answer

100

Answer

0.1

Question 4

The "Crude death rate" is defined as the number of deaths (from all causes) per 1000 estimated what?

Accepted Answer

Mid-year population

Answer

Total population

Answer

Total births

Answer

Live births

Question 5

Type 1 Sampling error is classified as?

Accepted Answer

Alpha error

Answer

Beta error

Answer

Gamma error

Answer

Delta error

Question 6

A village is divided into five relevant subgroups for the purpose of a survey. Individuals from each subgroup are then selected randomly. What is this type of sampling called?

Accepted Answer

Stratified sampling

Answer

Simple Random sampling

Answer

Cluster Sampling

Answer

Systematic Sampling

Question 7

A chi-square test would be most appropriate for testing which one of the following hypotheses?

Accepted Answer

That a smaller proportion of people who were immunized against chickenpox subsequently develop zoster than those who were not immunized.

Answer

That the mean score of students is greater than that of other students.

Answer

That the mean blood pressure of black and white male-hypertensive patients taking ACE inhibitors is the same as that of black and white female-hypertensive patients taking ACE inhibitors and that of black and white males and females taking diuretics and placebos.

Answer

That the mean cost of treating a patient with coronary artery disease with angioplasty is greater than the mean cost of providing medical treatment.

Question 8

What is true about the standard normal distribution?

Accepted Answer

The total area under the curve is equal to 1.

Answer

The mean is equal to 1.

Answer

The mean is greater than the median, which is greater than the mode.

Answer

The distribution has a tail towards the right.

Question 9

Which of the following is true about a cohort study?

Accepted Answer

Incidence can be calculated

Answer

It proceeds from effect to cause

Answer

It is inexpensive

Answer

It requires a shorter time than a case-control study

Question 10

Interpret the statistical graph shown below:

Accepted Answer

Positive correlation

Answer

Negative correlation

Answer

Absent correlation

Answer

Spurious correlation

Biostatistics — MCQs

Biostatistics — MCQs

On this page

Practice by Chapter

Want unlimited practice?