Biostatistics Practice Questions

Q: A new drug is to be evaluated for its therapeutic effect. The best study design will be.

Randomized controlled trial. ***Randomized controlled trial*** - This design is the **gold standard** for evaluating the effectiveness of a new therapeutic intervention. - **Randomization** minimizes confounding, and a control group allows for direct comparison of outcomes, isolating the **drug's effect**. *Natural experiment* - This design involves observing the effects of an intervention that occurs naturally, without researcher manipulation. - It lacks the **control** and **randomization** necessary to definitively attribute observed effects solely to the therapeutic agent. *Cross sectional survey* - This design assesses the prevalence of a condition or exposure at a single point in time. - It cannot establish **causality** or evaluate the therapeutic effect of an intervention over time. *Case control design* - This retrospective design compares individuals with a disease (cases) to those without (controls) to identify past exposures. - It is used to investigate **risk factors** for diseases, not to evaluate the therapeutic efficacy of a new drug.

Q: "Sampling error" occurs due to the variation in results

between one sample and another. ***between one sample and another*** - **Sampling error** arises because a sample is not a perfect representation of the entire population from which it is drawn. - This error quantifies the natural **variability** that occurs when different subgroups (samples) are selected from the same population. *due to the use of many instruments in the study* - This scenario describes **inter-instrument variability** or **measurement error**, which is related to the precision and calibration of different tools. - While it can introduce error, it is distinct from sampling error, which arises from the representativeness of the chosen study subjects. *due to the multiple readings taken on the same instrument* - Multiple readings on the same instrument assess **intra-instrument variability** or **repeatability**, indicating how consistent a single instrument is over time. - This relates to the precision of the measurement device, not the representativeness of the sample itself. *between the observations of two individuals* - Differences in observations between two individuals indicate **inter-rater variability** or **observer bias**. - This type of error is related to subjective interpretation or measurement technique by different observers, rather than the intrinsic variability between selected samples.

Q: Which is the fertility indicator that gives the approximate magnitude of completed family size?

Total Fertility Rate. ***Total Fertility Rate*** - The **Total Fertility Rate (TFR)** estimates the average number of children a woman would have over her lifetime if she were to experience current age-specific fertility rates. - It provides a measure of the **completed family size** in a hypothetical cohort of women. *Age-specific Fertility Rate* - The **Age-specific Fertility Rate (ASFR)** measures the number of births to women in a particular age group per 1,000 women in that age group. - It does not directly provide the completed family size but is a component used to calculate the TFR. *General Fertility Rate* - The **General Fertility Rate (GFR)** calculates the number of live births per 1,000 women of childbearing age (typically 15-49 years) in a given year. - While it reflects overall fertility, it does not provide an estimate of the completed family size per woman. *Gross Reproduction Rate* - The **Gross Reproduction Rate (GRR)** is similar to the TFR but only considers female births. - It estimates the average number of daughters a woman would have over her lifetime based on current age-specific fertility rates, without accounting for mortality.

Q: In a cohort study spanning 20 years, 50 out of 5000 smokers developed lung cancer, and 10 out of 10000 non-smokers developed lung cancer. What is the 'relative risk' among smokers for developing lung cancer?

10. ***Correct Answer: 10*** - The incidence of lung cancer in smokers = 50/5000 = 0.01 (1%) - The incidence of lung cancer in non-smokers = 10/10000 = 0.001 (0.1%) - **Relative Risk (RR) = Incidence in exposed / Incidence in unexposed** - RR = 0.01 / 0.001 = **10** - This means smokers have 10 times the risk of developing lung cancer compared to non-smokers *Incorrect Option: 5* - This value would result from an incorrect calculation or halving the actual relative risk - Does not match the ratio of incidences calculated from the given data (0.01/0.001 ≠ 5) - Would underestimate the true risk among smokers *Incorrect Option: 45* - This does not represent any standard epidemiological measure from this data - May result from confusion with absolute numbers or incorrect arithmetic - Neither the absolute risk difference nor any valid ratio yields this number *Incorrect Option: 50* - This represents the **absolute number of cases** in the smoking cohort, not a risk measure - Relative risk is a **ratio** comparing incidence rates between groups, not a count - Common error: confusing absolute numbers with relative measures

Q: Why is matching done in a case-control study?

To remove the effect of known confounders. ***To remove the effect of known confounders*** - **Matching** in a case-control study helps to control for the influence of specific variables (known confounders) that might otherwise distort the observed association between the exposure and outcome. - By matching cases and controls on characteristics like age, sex, or socioeconomic status, researchers ensure that these factors are similarly distributed in both groups, isolating the effect of the primary exposure. *To remove the effect of unknown confounders* - Matching is effective for **known confounders** that are identified and controlled during study design. - It does not address the impact of **unknown** or unmeasured confounding variables, which can still influence the study's results. *To eliminate selection bias* - **Selection bias** occurs when participants are not representative of the target population; matching primarily addresses confounding, not the initial selection process. - While careful selection is crucial to minimize selection bias, matching itself is more related to controlling for nuisance variables *after* selection. *To eliminate interviewer’s bias* - **Interviewer bias** (or observer bias) arises when the interviewer's expectations or knowledge influence data collection or interpretation. - This type of bias is typically addressed through blinding (e.g., blinded interviewers, participants, or outcome assessors), not through matching study participants.

Q: Retrospective cohort studies have the following features EXCEPT:

Generally more expensive than prospective studies. ***Generally more expensive than prospective studies*** - Retrospective cohort studies are typically **less expensive** than prospective studies because they utilize existing data, thus avoiding the costs associated with new data collection and long-term follow-up. - The primary expenses in retrospective studies often involve data retrieval and analysis, which are generally lower compared to the extensive resources needed for prospective data acquisition. *Investigator goes back in time to select study groups* - This is a hallmark feature of **retrospective cohort studies**, where researchers define study groups (exposed and unexposed) based on past exposures. - Data collection then proceeds by looking forward from the exposure to identify health outcomes that have already occurred. *Outcomes have occurred before the start of the study* - In a **retrospective cohort study**, both the exposure and the **outcomes of interest** have already taken place before the study officially begins. - Researchers identify existing records to link past exposures to these pre-existing outcomes. *Results are obtained more quickly* - Because all exposures and outcomes have already occurred and data is often readily available, **retrospective studies** can generate results much faster than prospective studies. - They do not require waiting for events to unfold in real-time, which significantly reduces the duration of the research process.

Q: In a normal distribution, Mean ± 2 S.D. contains

95.4 % values. ***95.4 % values*** - According to the **empirical rule** (or 68-95-99.7 rule) for normal distributions, approximately **95.4%** of data falls within two standard deviations of the mean. - This interval covers from (Mean - 2 S.D.) to (Mean + 2 S.D.) and represents the likelihood of a value falling in this range. *68.3 % values* - This percentage corresponds to the data contained within **Mean ± 1 S.D.** in a normal distribution, not Mean ± 2 S.D. - It signifies that roughly two-thirds of all observations lie within one standard deviation from the mean in a bell-shaped curve. *91.2 % values* - This value is not a standard percentage associated with common multiples of standard deviations (1, 2, or 3) from the mean in a normal distribution. - It does not correspond to any universally recognized interval like ±1 S.D., ±2 S.D., or ±3 S.D. *99.7 % values* - This percentage represents the data contained within **Mean ± 3 S.D.** in a normal distribution. - It indicates that almost all (99.7%) of the data points are expected to fall within three standard deviations from the mean.

Q: The appropriate statistical test to find out obesity as a significant risk factor for breast cancer is:

Chi-square test. ***Chi-square test*** - The **chi-square test** is used to determine if there is a **significant association** between two **categorical variables**. - In this scenario, both obesity (yes/no) and breast cancer (yes/no) are categorical, making chi-square appropriate to assess if obesity is a risk factor. *Wilcoxon’s signed rank test* - This is a **non-parametric test** used for comparing two related samples or repeated measurements on a single sample, especially when data are not normally distributed. - It is not suitable for assessing the association between two independent categorical variables like obesity and breast cancer. *Student’s paired ‘t’ test* - The **paired t-test** is used to compare the means of two related groups or measurements from the same subject under different conditions (e.g., before and after an intervention). - This test is designed for **continuous data** and would not be appropriate for the categorical variables of obesity and breast cancer. *Student’s unpaired ‘t’ test* - The **unpaired t-test** (also known as independent samples t-test) is used to compare the means of two independent groups for a **continuous outcome variable**. - It is not suitable when both the exposure (obesity) and the outcome (breast cancer) are categorical variables.

Question 1

A new drug is to be evaluated for its therapeutic effect. The best study design will be.

Accepted Answer

Randomized controlled trial

Answer

Natural experiment

Answer

Cross sectional survey

Answer

Case control design

Question 2

"Sampling error" occurs due to the variation in results

Accepted Answer

between one sample and another

Answer

due to the use of many instruments in the study

Answer

due to the multiple readings taken on the same instrument

Answer

between the observations of two individuals

Question 3

The "Relative Risk" of 0.25 indicates

Accepted Answer

75% reduction in the incidence rate in the exposed individuals compared with the unexposed

Answer

2.5 times higher risk in the exposed individuals compared with the unexposed

Answer

25% increase in the incidence rate in the exposed individuals compared with the unexposed

Answer

75% risk increase in the exposed individuals compared with the unexposed

Question 4

Which is the fertility indicator that gives the approximate magnitude of completed family size?

Accepted Answer

Total Fertility Rate

Answer

Age-specific Fertility Rate

Answer

General Fertility Rate

Answer

Gross Reproduction Rate

Question 5

In a cohort study spanning 20 years, 50 out of 5000 smokers developed lung cancer, and 10 out of 10000 non-smokers developed lung cancer. What is the 'relative risk' among smokers for developing lung cancer?

Accepted Answer

10

Answer

5

Answer

45

Answer

50

Question 6

Why is matching done in a case-control study?

Accepted Answer

To remove the effect of known confounders

Answer

To remove the effect of unknown confounders

Answer

To eliminate selection bias

Answer

To eliminate interviewer’s bias

Question 7

Retrospective cohort studies have the following features EXCEPT:

Accepted Answer

Generally more expensive than prospective studies

Answer

Investigator goes back in time to select study groups

Answer

Outcomes have occurred before the start of the study

Answer

Results are obtained more quickly

Question 8

In a normal distribution, Mean ± 2 S.D. contains

Accepted Answer

95.4 % values

Answer

68.3 % values

Answer

91.2 % values

Answer

99.7 % values

Question 9

In a case control study, confounding factors can be minimized by the following EXCEPT:

Accepted Answer

Increasing sample size for cases and controls

Answer

Matching of variables such as age and sex

Answer

Stratification during analysis

Answer

Randomization during selection

Question 10

The appropriate statistical test to find out obesity as a significant risk factor for breast cancer is:

Accepted Answer

Chi-square test

Answer

Wilcoxon’s signed rank test

Answer

Student’s paired ‘t’ test

Answer

Student’s unpaired ‘t’ test

Biostatistics — MCQs

Biostatistics — MCQs

On this page

Practice by Chapter

Want unlimited practice?