I85PREP
⌘ K
Arab Board Community Medicine Final Written study materials
Arab Board Community Medicine Final Written study resources

Arab Board Community Medicine Final Written practice questions

One question at a time. Build your Arab Board Community Medicine Final Written understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Epidemiology

In a closed cohort followed for one year, which measure is the cumulative incidence?

  1. New cases during the year divided by total person-years of follow-up
  2. All existing cases at the end of the year divided by the total population
  3. New cases in the year divided by the population at risk at the start Correct answer.
  4. Deaths during the year divided by all cases identified

Explanation

Cumulative incidence, or incidence proportion, is the number of new cases arising in a defined period divided by the number of people at risk at the beginning of that period. It estimates the average risk of developing disease in the cohort. A person-time denominator gives incidence density rather than cumulative incidence.

Question 2Epidemiology

Which change would increase the prevalence of a chronic disease while leaving its incidence unchanged?

  1. A rise in the cure rate
  2. Improved survival of affected patients Correct answer.
  3. A rise in the case-fatality rate
  4. A fall in the average duration of disease

Explanation

Prevalence is approximately incidence multiplied by average disease duration. Improved survival lengthens duration, so prevalence rises even when incidence is stable. Higher cure rates, higher fatality, and shorter duration all reduce prevalence.

Question 3Epidemiology

Sensitivity of a diagnostic test is defined as the proportion of people with disease who:

  1. test positive Correct answer.
  2. are correctly classified overall
  3. are free of disease and test negative
  4. test negative

Explanation

Sensitivity is the probability of a positive test among people who truly have the disease, calculated as true positives divided by true positives plus false negatives. It reflects the test's ability to detect disease. Specificity concerns people without disease.

Question 4Epidemiology

Specificity of a screening test refers to the proportion of people without the disease who:

  1. test negative Correct answer.
  2. test positive
  3. have a false-negative result
  4. are detected by the test

Explanation

Specificity is the probability of a negative test among people who do not have the disease, calculated as true negatives divided by true negatives plus false positives. A highly specific test produces few false-positive results. Sensitivity concerns people with disease.

Question 5Epidemiology

A screening test has sensitivity 90% and specificity 90%. In a population with prevalence 1%, what is the approximate positive predictive value?

  1. About 99%
  2. About 50%
  3. About 8% Correct answer.
  4. About 90%

Explanation

Positive predictive value equals sensitivity multiplied by prevalence, divided by that product plus the false-positive rate multiplied by one minus prevalence. With 1% prevalence, PPV is 0.009 divided by 0.108, which is about 8%. Low prevalence generates many false positives even when sensitivity and specificity are both good.

Question 6Epidemiology

Which measure of association is estimated in a case-control study?

  1. Relative risk
  2. Odds ratio Correct answer.
  3. Incidence rate ratio
  4. Attributable risk

Explanation

Case-control studies sample people by outcome status, so incidence cannot be measured directly. The odds ratio is calculated from exposure odds among cases and controls and approximates the risk ratio when the disease is rare. Relative risk and incidence rate ratio require cohort or follow-up data.

Question 7Epidemiology

Which measure is directly obtained in a cohort study comparing exposed and unexposed groups?

  1. Prevalence ratio only
  2. Relative risk Correct answer.
  3. Case-fatality rate
  4. Odds ratio only

Explanation

Cohort studies follow exposed and unexposed groups and measure incident disease, so the ratio of risks, or relative risk, can be calculated directly. Odds ratios are typical of case-control designs. Prevalence requires cross-sectional sampling, and case fatality concerns deaths among cases.

Question 8Epidemiology

What is the primary purpose of random allocation in a clinical trial?

  1. To ensure participants complete follow-up
  2. To increase the sample size
  3. To balance known and unknown confounders Correct answer.
  4. To mask treatment assignment from assessors

Explanation

Randomization makes treatment groups comparable on average for both measured and unmeasured confounders, which supports causal inference. Blinding addresses observer bias, and follow-up completeness concerns retention rather than allocation. Sample size is determined by power calculations.

Question 9Epidemiology

A variable is a confounder when it is:

  1. a consequence of the exposure and also a direct cause of the outcome
  2. linked to both exposure and outcome, not on the causal pathway Correct answer.
  3. measured with random error in both groups
  4. associated with the outcome only among the exposed

Explanation

A confounder distorts the exposure-outcome relation because it is associated with the exposure and independently with the outcome, and it does not lie on the causal pathway. A variable that is a consequence of exposure and a cause of outcome is a mediator, not a confounder. Random measurement error causes misclassification rather than confounding.

Question 10Epidemiology

Using hospital controls in a case-control study can introduce which bias when the exposure itself causes hospital admission?

  1. Interviewer bias
  2. Berkson's bias Correct answer.
  3. Lead-time bias
  4. Recall bias

Explanation

Hospital controls may overrepresent exposures that are linked to hospitalization, so their exposure prevalence may not reflect the source population. This is a form of selection bias often called Berkson's bias. Recall and interviewer bias concern how exposure information is collected.

Question 11Epidemiology

Cases with a disease recall past exposures more thoroughly than healthy controls. This is:

  1. Ecological fallacy
  2. Recall bias Correct answer.
  3. Confounding
  4. Selection bias

Explanation

Recall bias is differential accuracy of remembered exposure information between compared groups. It is an information bias that can distort effect estimates in case-control and retrospective studies. Selection bias concerns who enters the study, and confounding concerns a third variable.

Question 12Epidemiology

Screening detects a disease earlier and patients appear to survive longer, but the time of death is unchanged. This phenomenon is:

  1. Overdiagnosis
  2. Selection bias
  3. Length bias
  4. Lead-time bias Correct answer.

Explanation

Lead-time bias is the apparent survival gain created solely by advancing the time of diagnosis, not by delaying death. Survival measured from diagnosis lengthens even when the natural history is unchanged. Length bias concerns slower-growing cases, and overdiagnosis concerns detection of disease that would never cause harm.

Question 13Epidemiology

Screening programmes preferentially detect slowly progressive disease with a long preclinical phase. This is:

  1. Lead-time bias
  2. Confounding
  3. Length bias Correct answer.
  4. Recall bias

Explanation

Length bias, or length-time bias, arises because slow-growing disease spends more time detectable and is therefore more likely to be picked up by periodic screening. It makes screened cases appear to have a better prognosis. Lead-time bias is earlier diagnosis alone, and recall bias concerns exposure memory.

Question 14Epidemiology

In a cohort study, the attributable risk among the exposed is calculated as:

  1. Exposed incidence minus unexposed incidence Correct answer.
  2. Incidence in exposed divided by incidence in unexposed
  3. Incidence in total population minus incidence in unexposed
  4. Odds of exposure in cases divided by odds in controls

Explanation

Attributable risk, or risk difference, is the excess incidence among exposed people that can be attributed to the exposure. It is an absolute measure calculated by subtracting unexposed incidence from exposed incidence. The ratio of the two incidences is relative risk.

Question 15Epidemiology

The population attributable fraction represents the:

  1. Proportion of all disease prevented by eliminating exposure Correct answer.
  2. Number needed to treat to prevent one case
  3. Proportion of exposed cases caused by the exposure
  4. Risk of disease in the exposed relative to the unexposed

Explanation

Population attributable fraction combines the strength of association and the prevalence of exposure to estimate the fraction of all disease in the population attributable to that exposure. It is useful for public health priority setting. The exposed attributable fraction applies only to exposed cases, and relative risk is a ratio measure.

Question 16Epidemiology

In a trial, the event rate is 20% with placebo and 10% with treatment. What is the number needed to treat to prevent one event?

  1. 10 Correct answer.
  2. 5
  3. 2
  4. 20

Explanation

Absolute risk reduction is 20% minus 10%, which equals 10%, or 0.10. Number needed to treat is 1 divided by absolute risk reduction, so 1 divided by 0.10 equals 10. It expresses how many patients must be treated for one additional favourable outcome.

Question 17Epidemiology

Intention-to-treat analysis in a randomized trial primarily:

  1. adjusts for confounding after randomization
  2. increases the observed treatment effect by excluding non-adherent participants
  3. preserves the benefits of randomization and avoids attrition bias Correct answer.
  4. reduces the sample size needed for power

Explanation

Intention-to-treat analyses all participants in their randomized groups, preserving comparability and avoiding bias from selective dropout or non-adherence. It may dilute the estimated efficacy of actually receiving treatment. Per-protocol analysis can exaggerate effects and reintroduce confounding.

Question 18Epidemiology

Matching in a case-control study is used mainly to:

  1. control measured confounders by design Correct answer.
  2. allow direct calculation of incidence
  3. eliminate recall bias
  4. increase the precision of the odds ratio by increasing sample size

Explanation

Matching selects controls with similar values of potential confounders, such as age and sex, as the cases, so those factors cannot distort the comparison. It controls confounding at the design stage, although a matched analysis is still required. Matching does not remove information bias or permit incidence calculation.

Question 19Epidemiology

Direct age standardization is used when comparing mortality rates between two countries because:

  1. crude rates are always biased
  2. it adjusts for differences in disease severity
  3. the countries have different age structures Correct answer.
  4. standardization removes random error

Explanation

Crude mortality depends on both age-specific rates and population age composition. Direct standardization applies the study population's age-specific rates to a common standard population so comparisons are not distorted by age structure. It does not remove random error or adjust for disease severity.

Question 20Epidemiology

Which statement best distinguishes effect modification from confounding?

  1. Effect modification is real variation across strata; confounding is distortion to be adjusted Correct answer.
  2. Both phenomena are removed by stratification, but effect modification also changes the crude estimate in every study
  3. Effect modification is a nuisance to be removed by adjustment, whereas confounding is a genuine biological interaction between exposures
  4. Confounding is reported as stratum-specific estimates, whereas effect modification is pooled into one adjusted estimate

Explanation

Effect modification, or interaction, means the exposure effect genuinely differs across strata of a third variable, and stratum-specific effects should be reported. Confounding is a distortion of the exposure-outcome relation that should be controlled to obtain an adjusted estimate. Stratification can explore both, but their interpretation differs.

Question 21Epidemiology

A crude odds ratio is 2.0. After stratification, stratum-specific odds ratios are both 1.2 and the Mantel-Haenszel pooled estimate is 1.2. This pattern indicates:

  1. effect modification by the stratification variable
  2. a protective effect of the exposure
  3. random misclassification of exposure
  4. confounding by the stratification variable Correct answer.

Explanation

When stratum-specific estimates are similar to each other but differ from the crude estimate, the stratified variable is a confounder. The Mantel-Haenszel estimate pools the homogeneous strata to give an adjusted effect. Effect modification would show materially different stratum-specific estimates.

Question 22Epidemiology

A study finds that countries with higher average salt intake have higher stroke mortality. The error of concluding that individuals who eat more salt have higher stroke risk is:

  1. lead-time bias
  2. Berkson's bias
  3. confounding
  4. ecological fallacy Correct answer.

Explanation

The ecological fallacy is inferring individual-level associations from group-level data. Group averages can conceal within-group variation, and associations may differ at the individual level. Confounding and selection bias are separate threats to validity.

Question 23Epidemiology

In an outbreak investigation, which step comes immediately after confirming that an outbreak exists?

  1. Write the final report
  2. Implement control measures
  3. Perform an analytical study
  4. Establish a case definition Correct answer.

Explanation

After verifying the outbreak, investigators establish a case definition so that cases can be identified and counted consistently. Descriptive epidemiology and hypothesis generation follow, with analytical studies later if needed. Control measures may begin early, but the case definition is the immediate systematic step.

Question 24Epidemiology

For a disease with a basic reproduction number (R0) of 4, the approximate herd immunity threshold is:

  1. 50%
  2. 25%
  3. 75% Correct answer.
  4. 67%

Explanation

The herd immunity threshold is calculated as 1 minus 1 divided by R0. For R0 equal to 4, this is 1 minus 0.25, or 0.75, which is 75%. The threshold rises as transmissibility increases.

Question 25Epidemiology

The secondary attack rate measures the:

  1. number of new cases in the whole population per year
  2. proportion of susceptible contacts becoming ill after exposure Correct answer.
  3. proportion of all cases who die from the disease
  4. ratio of primary to secondary cases in an outbreak

Explanation

The secondary attack rate is the number of secondary cases among susceptible contacts divided by the number of susceptible contacts, within the relevant incubation period. It reflects infectiousness and transmission in close-contact settings. Case fatality and incidence rates are different measures.

Question 26Epidemiology

In a screening programme, all participants receive both the new screening test and a definitive diagnostic assessment. The proportion of people with the disease who are correctly classified by the screening test as having the disease is called:

  1. Prevalence
  2. Sensitivity Correct answer.
  3. Positive predictive value
  4. Specificity

Explanation

Sensitivity is the probability that a person who truly has the disease is classified as positive, calculated as true positives divided by true positives plus false negatives. It describes how well the test detects disease among those who actually have it. Specificity is conditional on being disease-free, whereas predictive values are conditional on the test result.

Question 27Epidemiology

Which measure of disease frequency explicitly incorporates the varying length of time that each person remains under observation?

  1. Incidence rate Correct answer.
  2. Cumulative incidence
  3. Period prevalence
  4. Point prevalence

Explanation

The incidence rate, or incidence density, divides the number of new cases by the total person-time at risk, so each person contributes the actual time observed. Cumulative incidence uses a fixed follow-up period with a denominator of persons at risk at the start and assumes complete follow-up. Prevalence measures existing cases rather than the occurrence of new cases per unit of time.

Question 28Epidemiology

A disease has a prevalence of 1% in a population. A screening test has a sensitivity of 90% and a specificity of 90%. Among 1000 people tested, the positive predictive value is approximately:

  1. About 8% Correct answer.
  2. About 90%
  3. About 47%
  4. About 50%

Explanation

In 1000 people with 1% prevalence there are 10 true cases, and at 90% sensitivity 9 of them are detected. Of the 990 people without disease, 10% (99) are falsely positive, giving 108 positive results and a positive predictive value of 9/108, or about 8%. When prevalence is low, even a fairly specific test produces many false positives relative to true positives.

Question 29Epidemiology

A cohort study reports a relative risk of 1.8 for a suspected exposure, with a 95% confidence interval of 0.9 to 3.6. Which interpretation is most appropriate?

  1. The association is significant, since 1.8 exceeds 1.0.
  2. No association cannot be excluded, since the interval includes 1.0. Correct answer.
  3. The association is significant, since the interval is narrow.
  4. There is no real association, since the interval is wide.

Explanation

A 95% confidence interval that includes the null value of 1.0 means the data are compatible with no association at the conventional 5% significance level, so the result is not statistically significant. The point estimate of 1.8 describes the observed effect size but does not by itself establish significance. A wide interval reflects imprecision rather than proof that no effect exists.

Question 30Epidemiology

In a case-control study, exposure is misclassified to a similar degree among cases and controls, and the errors are unrelated to disease status. The most likely effect on the odds ratio is:

  1. No change in the point estimate
  2. Bias away from the null value
  3. Bias toward the null value Correct answer.
  4. Systematic reversal of effect direction

Explanation

Non-differential misclassification of a binary exposure tends to dilute the contrast between cases and controls, shifting the odds ratio toward 1.0. It generally attenuates rather than exaggerates the observed association. Differential misclassification, in which errors differ by disease status, can bias in either direction.

Question 31Biostatistics and Research Methodology

Serum cholesterol values in a community survey are strongly skewed to the right by a small number of people with very high values. Which measure best summarises the typical value in this distribution?

  1. Range
  2. Mean
  3. Median Correct answer.
  4. Mode

Explanation

In a skewed distribution the mean is pulled towards the extreme tail, while the median depends only on the rank order of the observations. The median is therefore the preferred measure of central tendency for skewed data such as income or cholesterol. The mode and the range do not describe the centre of a skewed distribution adequately.

Question 32Biostatistics and Research Methodology

A new rapid test for dengue is compared with a reference standard. Which proportion defines the sensitivity of the rapid test?

  1. The proportion of people testing negative who are disease-free
  2. The proportion of people without disease who test negative
  3. The proportion of people testing positive who have disease
  4. The proportion of people with disease who test positive Correct answer.

Explanation

Sensitivity is the conditional probability of a positive test given that disease is present, calculated as true positives divided by all people with disease. It describes how well the test detects cases and is independent of prevalence. The other proportions describe specificity, positive predictive value and negative predictive value respectively.

Question 33Biostatistics and Research Methodology

In hypothesis testing, the probability of rejecting the null hypothesis when it is in fact true is denoted by:

  1. The p-value
  2. Power
  3. Alpha Correct answer.
  4. Beta

Explanation

Alpha is the pre-specified significance level and equals the probability of a type I error, that is, rejecting a true null hypothesis. Beta is the probability of a type II error and power equals one minus beta. The p-value is the probability of data at least as extreme as those observed, calculated assuming the null hypothesis is true.

Question 34Biostatistics and Research Methodology

Which measure best describes the total burden of diabetes in a community at a fixed point in time?

  1. Attack rate
  2. Case fatality rate
  3. Prevalence Correct answer.
  4. Incidence rate

Explanation

Prevalence counts all existing cases, both old and new, in a defined population at a given time and therefore reflects the overall burden of disease. Incidence measures only new cases arising over a period. Case fatality describes the proportion of cases that die and does not capture total burden.

Question 35Biostatistics and Research Methodology

A researcher wishes to compare the variability of birth weight measured in grams with the variability of gestational age measured in weeks. Which statistic is most appropriate?

  1. Standard deviation of the sample mean
  2. Coefficient of variation Correct answer.
  3. Interquartile range
  4. Variance

Explanation

The coefficient of variation expresses the standard deviation as a percentage of the mean, which removes the influence of the measurement unit. It is therefore the correct choice when comparing dispersion across variables measured on different scales. Variance and interquartile range remain expressed in the original units.

Question 36Biostatistics and Research Methodology

Which of the following variables is measured on a nominal scale?

  1. Pain severity rated as mild, moderate or severe
  2. Number of antenatal visits per woman
  3. Birth weight recorded in grams
  4. Blood group classified as A, B, AB or O Correct answer.

Explanation

A nominal scale assigns categories that have no inherent order or numerical distance, and blood group is a classic example. Ordinal scales such as pain severity have a meaningful rank, while weight and visit counts are measured on interval or ratio scales.

Question 37Biostatistics and Research Methodology

In a randomised controlled trial, the main purpose of random allocation of participants is to:

  1. Increase the total sample size available for analysis
  2. Balance known and unknown confounders between groups Correct answer.
  3. Remove laboratory measurement error from outcome assessment
  4. Make the sample representative of all patients in the population

Explanation

Randomisation gives every participant a known chance of allocation, so that prognostic factors, both measured and unmeasured, are distributed similarly across groups on average. This protects the internal validity of the comparison. Representativeness of the wider population is achieved through sampling rather than allocation.

Question 38Biostatistics and Research Methodology

In a placebo-controlled trial, mortality was 20% in the placebo group and 15% in the treated group. The number needed to treat to prevent one death is:

  1. 25
  2. 15
  3. 5
  4. 20 Correct answer.

Explanation

The absolute risk reduction is 20% minus 15%, or 0.05. The number needed to treat is the reciprocal of the absolute risk reduction, giving 1 divided by 0.05, which equals 20. The relative risk reduction of 25% is a separate quantity and does not substitute for the absolute risk reduction.

Question 39Biostatistics and Research Methodology

A screening test has a sensitivity of 90% and a specificity of 95%. When it is applied in a population where the prevalence of the disease is only 0.5%, the positive predictive value will most likely be:

  1. Low, because most positive results will be false positives Correct answer.
  2. Equal to the sensitivity of the test
  3. High, because a specificity of 95% rules out nearly all healthy people
  4. Unchanged, because sensitivity and specificity do not vary

Explanation

Predictive values depend on prevalence as well as on sensitivity and specificity. When prevalence is very low, the large number of people without disease generates many false positives even at 95% specificity, so a positive result is more likely to be false than true. Sensitivity and specificity themselves remain constant across populations, but the positive predictive value falls.

Question 40Biostatistics and Research Methodology

A 95% confidence interval for a mean is calculated from 100 observations. If the sample size is increased to 400 while the standard deviation stays the same, the interval will approximately:

  1. Leave the width unchanged but shift the estimate
  2. Double the width of the 95% confidence interval
  3. Halve the width of the 95% confidence interval Correct answer.
  4. Reduce the standard error by a factor of four

Explanation

The standard error of a mean is the standard deviation divided by the square root of the sample size. Quadrupling the sample size halves the standard error and therefore halves the width of the confidence interval. The standard error itself falls by a factor of two, not four.

Question 41Biostatistics and Research Methodology

A case-control study of lung cancer and smoking reports an odds ratio of 8.0 with a 95% confidence interval of 5.0 to 12.0. Which statement is the correct interpretation?

  1. The risk of lung cancer among smokers is eight times the risk among non-smokers
  2. Smoking causes eight extra cases of lung cancer per 100 people
  3. Eight per cent of lung cancer cases are attributable to smoking
  4. The odds of exposure were eight times higher among cases than among controls Correct answer.

Explanation

In a case-control study participants are sampled on the basis of outcome, so the odds ratio is calculated from the odds of exposure in cases compared with controls. Disease risk cannot be estimated directly because the ratio of cases to controls is fixed by the investigator. The confidence interval excludes one, indicating a statistically significant positive association.

Question 42Biostatistics and Research Methodology

A cohort study finds a higher incidence of lung cancer among coffee drinkers. Smokers are more likely to drink coffee and are also at higher risk of lung cancer, and smoking does not lie on the causal pathway between coffee and lung cancer. In this study smoking acts as:

  1. A confounder Correct answer.
  2. An effect modifier
  3. A mediator
  4. A source of selection bias

Explanation

A confounder is associated with both the exposure and the outcome and is not an intermediate step in the causal pathway. Smoking meets all three criteria, so the crude association between coffee and lung cancer is distorted. Such confounding can be controlled by stratification or regression adjustment.

Question 43Biostatistics and Research Methodology

A trial is planned with 80% power at a two-sided alpha of 0.05. Which statement is correct?

  1. There is a 20% probability of a type II error if the assumed effect size is true Correct answer.
  2. There is an 80% probability that the alternative hypothesis is false
  3. There is an 80% probability that a significant result would be due to chance
  4. There is a 20% probability of a type I error

Explanation

Power is the probability of rejecting the null hypothesis when the specified alternative is true, so one minus power equals beta, the type II error rate. With 80% power, beta is 20% for the effect size assumed in the calculation. The type I error rate remains fixed at the chosen alpha of 0.05.

Question 44Biostatistics and Research Methodology

Blood pressure is measured in 40 hypertensive patients before and after four weeks of treatment. Which test is most appropriate for comparing the mean values?

  1. One-way analysis of variance
  2. Independent samples t-test
  3. Paired t-test Correct answer.
  4. Chi-square test

Explanation

The two measurements come from the same individuals, so the observations are dependent and must be analysed as matched pairs. The paired t-test compares the mean within-person difference with zero. The independent samples t-test would ignore the correlation between the two readings.

Question 45Biostatistics and Research Methodology

A researcher compares the proportion of current smokers across three age groups in a cross-sectional survey. Which test is most appropriate?

  1. Chi-square test Correct answer.
  2. Pearson correlation
  3. One-way ANOVA
  4. Paired t-test

Explanation

Both the exposure grouping and the smoking status are categorical, so the comparison concerns proportions in a contingency table. The chi-square test of independence assesses whether smoking status is associated with age group. Tests of means, such as ANOVA or the t-test, require a continuous outcome.

Question 46Biostatistics and Research Methodology

In a randomised trial with incomplete adherence, the primary purpose of intention-to-treat analysis is:

  1. To preserve the balance created by randomization Correct answer.
  2. To shorten the follow-up period required
  3. To estimate the effect only in fully adherent patients
  4. To increase the number of measurements available

Explanation

Intention-to-treat analysis compares participants in the groups to which they were originally allocated, regardless of adherence. This maintains the prognostic balance produced by randomisation and protects against attrition bias. Excluding non-adherent participants tends to exaggerate treatment effects because adherence is related to prognosis.

Question 47Biostatistics and Research Methodology

A test for diabetes has a specificity of 95%. This means that the test will:

  1. Provide a positive predictive value of 95% in every population
  2. Detect 5% of all true cases of disease
  3. Correctly classify 95% of people with disease as positive
  4. Correctly classify 95% of people without disease as negative Correct answer.

Explanation

Specificity is the proportion of people who do not have the disease and who are correctly identified as negative by the test. A specificity of 95% implies a false-positive rate of 5% among healthy people. It says nothing about how many true cases are detected or about predictive values.

Question 48Biostatistics and Research Methodology

A researcher wishes to detect a smaller difference between two group means while keeping alpha and power unchanged. The required sample size will:

  1. Decrease
  2. Remain the same
  3. Increase Correct answer.
  4. Halve

Explanation

Sample size is inversely related to the square of the difference that must be detected. A smaller true difference is harder to distinguish from random variation, so more participants are needed to maintain the same power. Alpha and power themselves determine the critical values and do not change here.

Question 49Biostatistics and Research Methodology

Which measure of association can be estimated directly from a cohort study but not from a case-control study?

  1. Correlation coefficient
  2. Odds ratio
  3. Relative risk Correct answer.
  4. Chi-square statistic

Explanation

A cohort study follows exposed and unexposed people and measures incidence in each group, so the ratio of these incidences, the relative risk, can be calculated directly. In a case-control study participants are selected by outcome, so incidence is unknown and only the odds ratio can be estimated. The odds ratio from a case-control study approximates the relative risk only when the outcome is rare.

Question 50Biostatistics and Research Methodology

In a community, 30% of adults smoke. The relative risk of lung cancer for smokers compared with non-smokers is 4. Using the Levin formula, the population attributable fraction is closest to:

  1. 25%
  2. 75%
  3. 30%
  4. 47% Correct answer.

Explanation

The Levin formula gives the population attributable fraction as Pe(RR minus 1) divided by 1 plus Pe(RR minus 1), where Pe is the prevalence of exposure. Substituting 0.30 and 4 gives 0.90 divided by 1.90, which is approximately 0.47. This represents the proportion of cases in the population that would be avoided if the exposure were eliminated.