MPH COMPREHENSIVE EXAM REVIEW
EXAM WITH CORRECT ACTUAL QUESTIONS
AND CORRECTLY WELL DEFINED ANSWERS
LATEST ALREADY GRADED A+
A public heath scientist wanted to calculate the linear association between low
density lipoprotein (LDL) and prostate specific antigen. The two variables are
continuous but the data are not normally distributed. Which of the following
statistical methods may be most appropriate?
Simple linear regression
Pearson's correlation coefficient
Cox proportional hazards model
Poisson regression
Spearman correlation coefficient - ANSWERS-Spearman correlation coefficient
Chlamydia trachomatis (CT) is an established pathogenic microbe in non-
gonococcal urethritis, pelvic inflammatory disease, and other infectious
diseases. An investigator wishes to conduct a study to examine the risk factors
associated with CT using 862 female undergraduate students. He or she intends
to control for several risk factors simultaneously in the analysis of variables
associated with CT. If CT is measured in binary scale, which of the following
analytic methods might be appropriate?
,Simple linear regression
Conditional multivariable logistic regression
Stratifies analysis
Multivariable logistic regression
Multivariate logistic regression - ANSWERS-Multivariable logistic regression
An investigator wishes to examine the relation between cholesterol as a risk for
heart disease. If he/she chooses to measure cholesterol using an interval scale,
which of the following scales will be adequate in measuring heart disease
(outcome variable)?
Ordinal
Continuous
Quantitative
Interval
Binary - ANSWERS-Binary
African-American race/ethnicity is a known risk factor for hypertension. A study
was conducted to determine the association between hypertension and salt
intake in a sample of 250 white and 250 black males. The investigators utilized
logistic regression model for their analysis in order to generate the coefficients
(β) to relate hypertension to salt intake (X1) and hypertension to race (X2).
What is the accurate interpretation of β1?
Log odds of hypertension associated with salt intake
Log odds of hypertension among the unexposed
,None of the above
Change in the mean hypertension with a unit change is salt intake given that
race is held at constant
Log odds of the hypertension associated with race - ANSWERS-log odds of
hypertension associated with salt intake
Intervention at individual and group levels have been conducted to reduce HIV
risk behaviors among the most vulnerable populations at highest risk in the
United States. In spite of these interventions, African-American women remain
the fastest growing segment of incident AIDS cases. Obviously, new research
studies are needed to provide intervention at individual, family, community,
and society levels. As an HIV/AIDS interventionist, which of the following
factors will you consider as HIV/AIDS-related behavior to incorporate into your
model in reducing HIV risky behaviors?
Subjective norms
Self-efficacy
Knowledge
Perceived risk
All of the above - ANSWERS-All of the above
Obesity intervention often utilized community trials to increase physical
activity. These trials are usually part of a "healthy heart" program and
incorporate social and environmental measures, including public service
, announcements about physical activity and total well-being. This approach in
terms of health behavior model is best described as the:
Social support model
Ecologic model
Health belief model
Transtheoretical model
Information, motivation and
skill building model - ANSWERS-Ecologic model
Multiple regression procedures may be used when two or more interval-level
measures serve as predictors of some normally distributed interval-level
dependent variable. In this model, the regression coefficient for any particular
independent or explanatory variable (X1) represents the change in the response
or dependent variable (Y) associated with a one-unit change in X1, while
controlling for or maintaining other predictors (X2, X3, etc.) at constant. If
required to use this predictive model in the analysis of study data, which of the
following assumptions are NOT necessary for a valid result?
All variables are interval- or ratio-scaled
The dependent variable should be normally distributed around the prediction
line
The independent variable should be normally distributed around the prediction
line
None of the above - ANSWERS-The independent variable should be normally
distributed around the prediction line
EXAM WITH CORRECT ACTUAL QUESTIONS
AND CORRECTLY WELL DEFINED ANSWERS
LATEST ALREADY GRADED A+
A public heath scientist wanted to calculate the linear association between low
density lipoprotein (LDL) and prostate specific antigen. The two variables are
continuous but the data are not normally distributed. Which of the following
statistical methods may be most appropriate?
Simple linear regression
Pearson's correlation coefficient
Cox proportional hazards model
Poisson regression
Spearman correlation coefficient - ANSWERS-Spearman correlation coefficient
Chlamydia trachomatis (CT) is an established pathogenic microbe in non-
gonococcal urethritis, pelvic inflammatory disease, and other infectious
diseases. An investigator wishes to conduct a study to examine the risk factors
associated with CT using 862 female undergraduate students. He or she intends
to control for several risk factors simultaneously in the analysis of variables
associated with CT. If CT is measured in binary scale, which of the following
analytic methods might be appropriate?
,Simple linear regression
Conditional multivariable logistic regression
Stratifies analysis
Multivariable logistic regression
Multivariate logistic regression - ANSWERS-Multivariable logistic regression
An investigator wishes to examine the relation between cholesterol as a risk for
heart disease. If he/she chooses to measure cholesterol using an interval scale,
which of the following scales will be adequate in measuring heart disease
(outcome variable)?
Ordinal
Continuous
Quantitative
Interval
Binary - ANSWERS-Binary
African-American race/ethnicity is a known risk factor for hypertension. A study
was conducted to determine the association between hypertension and salt
intake in a sample of 250 white and 250 black males. The investigators utilized
logistic regression model for their analysis in order to generate the coefficients
(β) to relate hypertension to salt intake (X1) and hypertension to race (X2).
What is the accurate interpretation of β1?
Log odds of hypertension associated with salt intake
Log odds of hypertension among the unexposed
,None of the above
Change in the mean hypertension with a unit change is salt intake given that
race is held at constant
Log odds of the hypertension associated with race - ANSWERS-log odds of
hypertension associated with salt intake
Intervention at individual and group levels have been conducted to reduce HIV
risk behaviors among the most vulnerable populations at highest risk in the
United States. In spite of these interventions, African-American women remain
the fastest growing segment of incident AIDS cases. Obviously, new research
studies are needed to provide intervention at individual, family, community,
and society levels. As an HIV/AIDS interventionist, which of the following
factors will you consider as HIV/AIDS-related behavior to incorporate into your
model in reducing HIV risky behaviors?
Subjective norms
Self-efficacy
Knowledge
Perceived risk
All of the above - ANSWERS-All of the above
Obesity intervention often utilized community trials to increase physical
activity. These trials are usually part of a "healthy heart" program and
incorporate social and environmental measures, including public service
, announcements about physical activity and total well-being. This approach in
terms of health behavior model is best described as the:
Social support model
Ecologic model
Health belief model
Transtheoretical model
Information, motivation and
skill building model - ANSWERS-Ecologic model
Multiple regression procedures may be used when two or more interval-level
measures serve as predictors of some normally distributed interval-level
dependent variable. In this model, the regression coefficient for any particular
independent or explanatory variable (X1) represents the change in the response
or dependent variable (Y) associated with a one-unit change in X1, while
controlling for or maintaining other predictors (X2, X3, etc.) at constant. If
required to use this predictive model in the analysis of study data, which of the
following assumptions are NOT necessary for a valid result?
All variables are interval- or ratio-scaled
The dependent variable should be normally distributed around the prediction
line
The independent variable should be normally distributed around the prediction
line
None of the above - ANSWERS-The independent variable should be normally
distributed around the prediction line
