MPH516 Module 1 Quiz Latest 2021 August

MPH516 Epidemiology and Biostatistics for Public Health Application 2

Module 1 Quiz

Question 1Assuming that the following sample table is for a cohort study, define the risk difference or attributable risk:

  (A/A + C) / (B/B + D)

  (A/A + B) / (C/C + D)

  (A/A + C) − (B/B + D)

  (A/A + B) − (C/C + D)

  None of these is correct.

SAMPLE 2 BY 2 TABLE

 Outcome

Total

Factor

+

−

 +

A

B

A + B

−

C

D

C + D

Total

A + C

B + D

A + B + C + D

 Question 2For which regression assumption does the Durbin–Watson statistic test?

  Linearity

  Independence of errors

  Homoscedasticity

  Multicollinearity

Question 3If it is accepted that an observed association is a causal one, an estimate of the impact that a successful preventive program might have can be derived from:

  relative risk.

  higher life expectancy.

  attributable risk.

  prevalence rates.

  All are correct.

 Question 4The death rate per 100,000 for lung cancer is 7 among nonsmokers and 71 among smokers. The death rate per 100,000 for coronary thrombosis is 422 among nonsmokers and 599 among smokers. The prevalence of smoking in the population is 55%. The relative risk of dying for a smoker compared to a nonsmoker is:

  9.1 for lung cancer and 0.30 for coronary thrombosis.

  9.1 for lung cancer and 1.4 for coronary thrombosis.

  10.1 for lung cancer and 8.4 for coronary thrombosis.

  10.1 for lung cancer and 1.4 for coronary thrombosis.

  12.4 for lung cancer and 1.7 for coronary thrombosis.

 Question 5The death rate per 100,000 for lung cancer is 7 among nonsmokers and 71 among smokers. The death rate per 100,000 for coronary thrombosis is 422 among nonsmokers and 599 among smokers. The prevalence of smoking in the population is 55%. Among smokers, the etiologic fraction of disease due to smoking is:

  0.90 for lung cancer and 0.88 for coronary thrombosis.

  0.90 for lung cancer and 0.29 for coronary thrombosis.

  0.89 for lung cancer and 0.88 for coronary thrombosis.

  0.89 for lung cancer and 0.29 for coronary thrombosis.

  It cannot be determined from the information provided.

 Question 6The death rate per 100,000 for lung cancer is 7 among nonsmokers and 71 among smokers. The death rate per 100,000 for coronary thrombosis is 422 among nonsmokers and 599 among smokers. The prevalence of smoking in the population is 55%. The population etiologic fraction of disease due to smoking is:

  0.80 for lung cancer and 0.28 for coronary thrombosis.

  0.80 for lung cancer and 0.18 for coronary thrombosis.

  0.83 for lung cancer and 0.28 for coronary thrombosis.

  0.83 for lung cancer and 0.18 for coronary thrombosis.

  It cannot be determined from the information provided.

 

Question 7The death rate per 100,000 for lung cancer is 7 among nonsmokers and 71 among smokers. The death rate per 100,000 for coronary thrombosis is 422 among nonsmokers and 599 among smokers. The prevalence of smoking in the population is 55%. On the basis of the relative risk and etiologic fractions associated with smoking for lung cancer and coronary thrombosis, which of the following statements is most likely to be correct?

  Smoking seems much more likely to be causally related to coronary thrombosis than to lung cancer.

  Smoking seems much more likely to be causally related to lung cancer than to coronary thrombosis.

  Smoking seems to be equally causally related to lung cancer and coronary thrombosis.

  Smoking does not seem to be causally related to either lung cancer or coronary thrombosis.

  No comparative statement is possible between smoking and lung cancer or coronary thrombosis.

 Question 8The method of least squares is used to find out which of the following?

  The relationship of the gradient of the line

  The intercept of the line

  The gradient of the line

  The line of best fit

 Question 9The population etiologic fraction is a measure of the proportion of the disease rate in a population attributable to the exposure of interest. This measure of effect is influenced by:

  the relative risk of the disease in exposed individuals versus unexposed individuals.

  the prevalence of the disease in the population.

  the prevalence of the exposure in the population.

  the relative risk of the disease in exposed individuals versus unexposed individuals and the prevalence of the disease in the population.

  the relative risk of the disease in exposed individuals versus unexposed individuals and the prevalence of the exposure in the population.

 Question 10The population etiologic fraction for a particular disease from Factor X alone is five times greater than that from Factor Y alone. If the relative risk associated with Factor X is 2, and with Factor Y is 20, which of the following statements is true?

  The risk of developing the disease is greater in those exposed to Factor X only than in those exposed to Factor Y only.

  Fewer persons are exposed to Factor Y than to Factor X.

  The proportion of the disease in the population attributable to Factor Y is greater than that attributable to Factor X.

  More persons are exposed to Factor Y than to Factor X.

  The risk of developing the disease for persons exposed to Factor Y is five times greater than for persons exposed to Factor X.

 Question 11What does the following graph show?

  Heteroscedasticity

  Non-linearity

  Heteroscedasticity and non-linearity

  Regression assumptions that have been met

 Question 12What does the F-test measure?

  How much the model has improved the prediction of the outcome compared to the level of inaccuracy of the model

  How much the model is influenced by the predictors compared to the numbers of residuals

  How much the model relies on residual mean squares compared to the inaccuracies of the predictors

  None of the above

Question 13What does the t-statistic test?

  The null hypothesis that the value of b

(correlation coefficient) is negative

  The null hypothesis that the value of b

(correlation coefficient) is equal to 0

  The null hypothesis that the value of b

(correlation coefficient) is equal to 1

  The null hypothesis that the value of b

(correlation coefficient) is positive

 Question 14What is multicollinearity?

  When predictor variables correlate very highly with each other.

  When predictor variables have a linear relationship with the outcome variable.

  When predictor variables are correlated with variables not in the regression model.

  When predictor variables are independent.

 Question 15What is R2?

  The percentage of variance in the predictor accounted for by the outcome variable

  The proportion of variance in the outcome accounted for by the predictor variable or variables

  The proportion of variance in the predictor accounted for by the outcome variable

  The percentage of variance in the outcome accounted for by the predictor variable or variables

 Question 16When assessing a positive relationship between alcohol consumption and oral cancer using a case-control study, increasing the sample size of the study will result in which of the following?

  A lower p value and a smaller 95% confidence interval

  A greater odds ratio and a higher disease prevalence

  A lower p value, a smaller 95% confidence interval, and a greater odds ratio

  A lower p value, a smaller 95% confidence interval, a higher disease prevalence, and a greater odds ratio

  None of these is correct.

 Question 17Which of the following statements about the t-statistic in regression is not true?

  The t-statistic tests whether the regression coefficient, b, is equal to 0.

  The t-statistic provides some idea of how well a predictor predicts the outcome variable.

  The t-statistic can be used to see whether a predictor variable makes a statistically significant contribution to the regression model.

  The t-statistic is equal to the regression coefficient divided by its standard deviation.

 Question 18Which of the following statements about the F-ratio is true?

  The F-ratio is the ratio of variance explained by the model to the error in the model.

  The F-ratio is the ratio of variance explained by the model to the total variance in the outcome variable.

  The F-ratio is the ratio of error variance to the total variance.

  The F-ratio is the proportion of variance explained by the regression model.

 Question 19Which of the following statements about outliers is not true?

  Outliers are values very different from the rest of the data.

  Outliers bias the mean.

  Outliers bias regression parameters.

  Outliers are influential cases.

 Question 20Which of the following is not a reason why multicollinearity a problem in regression?

  It limits the size of R.

  It makes it difficult to assess the importance of individual predictors.

  It leads to unstable regression coefficients.

  It creates heteroscedasticity in the data.