CHLH 274- Final Exam Study Guide
Questions and Correct Answers.
What is the measure for an Ecological study - Answer correlation coefficient (R)
What is the measure for a Cross-Sectional study - Answer prevalence
What is the measure for a Case Control study - Answer odds ratio
What is the measure for a Cohort study - Answer relative risk
What is the measure for an Experimental study - Answer relative risk
In which 2 studies do you look exposure and disease at the same point in time? - Answer
ecological & cross sectional
In which study design do you start with the disease and trace back to the find the potential
exposure - Answer case control
in which study design do you start with exposure and and look for a potential outcome? -
Answer Cohort
in which study design do you ASSIGN exposure and look for potential outcomes - Answer
experimental
What three things do you need to eliminate for your study to be valid - Answer bias,
confounding, & random error
Who typically introduces bias? - Answer investigators or study participants
What are the 2 major types of bias? - Answer selection and observation.
what is selection bias - Answer selecting people to favor a certain outcomes (loss to follow
up can be cause)
what are the 2 types of selection bias - Answer Berkson & Neyman
, What are the 2 types of observation bias? - Answer recall & interviewer
what is confounding - Answer alternate explanation for association between exposure and
outcome (third variable)
When do you know there is confounding - Answer occurs when crude and adjusted
measures of effects are not equal (difference of at least 10%)
What are the 3 most common confounders - Answer age, gender, risk factors, (also race)
What is the best way to deal with confounding - Answer control it
describe how to control for confounding in the design phase - Answer major use of
restriction (restrict admissibility criteria), randomization, matching (get controls and
experimental to be as close as possible)
describe how to control for confounding in the analysis phase - Answer stratification of data
by age or gender, multivariate analysis
what is a random error - Answer not a systemic error, but a chance or luck of the draw
When do you reject your null hypothesis - Answer when p < alpha (0.05) (results are
significant not due to chance)
when do you fail to reject your null hypothesis - Answer when p> alpha (0.05) results are not
significant and are most likely due to chance
what two things does p reflect - Answer magnitude of association and sample size
What can you assume if 1 is within the confidence interval? - Answer that an error has
occurred and your results are not significantly significantly
Why are causal relationships important to public health - Answer they make public health
decisions and help design interventions
Questions and Correct Answers.
What is the measure for an Ecological study - Answer correlation coefficient (R)
What is the measure for a Cross-Sectional study - Answer prevalence
What is the measure for a Case Control study - Answer odds ratio
What is the measure for a Cohort study - Answer relative risk
What is the measure for an Experimental study - Answer relative risk
In which 2 studies do you look exposure and disease at the same point in time? - Answer
ecological & cross sectional
In which study design do you start with the disease and trace back to the find the potential
exposure - Answer case control
in which study design do you start with exposure and and look for a potential outcome? -
Answer Cohort
in which study design do you ASSIGN exposure and look for potential outcomes - Answer
experimental
What three things do you need to eliminate for your study to be valid - Answer bias,
confounding, & random error
Who typically introduces bias? - Answer investigators or study participants
What are the 2 major types of bias? - Answer selection and observation.
what is selection bias - Answer selecting people to favor a certain outcomes (loss to follow
up can be cause)
what are the 2 types of selection bias - Answer Berkson & Neyman
, What are the 2 types of observation bias? - Answer recall & interviewer
what is confounding - Answer alternate explanation for association between exposure and
outcome (third variable)
When do you know there is confounding - Answer occurs when crude and adjusted
measures of effects are not equal (difference of at least 10%)
What are the 3 most common confounders - Answer age, gender, risk factors, (also race)
What is the best way to deal with confounding - Answer control it
describe how to control for confounding in the design phase - Answer major use of
restriction (restrict admissibility criteria), randomization, matching (get controls and
experimental to be as close as possible)
describe how to control for confounding in the analysis phase - Answer stratification of data
by age or gender, multivariate analysis
what is a random error - Answer not a systemic error, but a chance or luck of the draw
When do you reject your null hypothesis - Answer when p < alpha (0.05) (results are
significant not due to chance)
when do you fail to reject your null hypothesis - Answer when p> alpha (0.05) results are not
significant and are most likely due to chance
what two things does p reflect - Answer magnitude of association and sample size
What can you assume if 1 is within the confidence interval? - Answer that an error has
occurred and your results are not significantly significantly
Why are causal relationships important to public health - Answer they make public health
decisions and help design interventions
