Deck 12: Choosing Designs for Effect Evaluations

Question 1

What type of study design, which comes from the field of analytic epidemiology, was initially developed for situations in which exposures are not manipulated, but assessed as they naturally occur?

A) Experimental designs
B) Quasi-experimental designs
C) Observational designs
D) Longitudinal designs

Answer 1

C

Question 2

What type of study design, when used with random assignment, is considered the strongest of evaluation designs?

A) One-group pretest/posttest
B) One-group posttest only
C) Two-group pretest/posttest
D) Two-group posttest only

Answer 2

C

Question 3

Which of the following occurs when the process of being involved in providing the pretest data in some way affects the posttest data?

A) Testing effect
B) Selection bias
C) Instrumentation
D) Manipulation

Answer 3

A

Question 4

For population-based evaluations, it is highly desirable to have what type of data on members of the population to determine if an intervention effect occurred?

A) Ordinal data
B) Baseline data
C) Aggregate data
D) Qualitative data

Question 5

What type of study design is generally used when the outcome is bounded?

A) Retrospective case-control design
B) Two-group prospective design
C) Multiple-group time-series design
D) Ecological design

Question 6

Which of the following refers to the loss of participants over time due to their dropping out of the program and/or evaluation, moving away, being lost to follow-up, or death?

A) Attrition
B) Manipulation
C) Maturation
D) Instrumentation

Question 7

What type of design considers the differences between groups rather than individuals and is used most often to determine whether there is an association between an exposure and a health consequence at the population level?

A) Retrospective case-control design
B) Two-group prospective design
C) Multiple-group time-series design
D) Ecological design

Question 8

When epidemiologic designs are used for program evaluation, they are no longer purely observational.

Question 9

In general, designs with fewer flaws are more complex and costly and are usually more likely to demonstrate a causal relationship between the health program and outcomes.

Question 10

Evaluation questions that ask whether the intervention changes the unbounded outcome at the individual level are best answered through observational designs.