Unit 06: Sampling Flashcards
What is a census in epidemiologic studies?
A method that systematically acquires and records relevant information about every member of a population.
Why is a census often impractical in epidemiology?
Time-consuming (contacting everyone in the population).
Costly (require large teams and resources).
Tedious (need a complete population listing).
Difficult if participation is voluntary.
Why is sampling often preferred over a census?
Sampling overcomes limitations of censuses by being:
Less time-consuming.
Less costly.
More practical for large populations.
The population to which it might be possible to extrapolate results is called the ____ population
target
A list of all study subjects in the source population, including their contact information, is called the _______.
sampling frame
What is the difference between the source population and the study/sampled population?
The source population is the group from which the sample is drawn, while the study/sampled population is the subset of sampling units that were actually selected for the study.
(True/False): The target population represents the group we actually study
False (the target population is the group we want our results to apply to, not necessarily the group we study).
(True/False): The sampling frame must include every member of the source population to ensure accurate results.
True
The ____ is the actual unit of measurement, such as a person, animal, household, or farm.
sampling unit
Sampling units are drawn from the ____to form the ____ population.
sampling frame, study/sampled
Why is it important to clearly define the population of interest?
A) To narrow down participants to save time and resources
B) To ensure results are representative of the population of interest
C) To identify the correct sampling frame and sampling units
D) All of the above
D
____ validity refers to the degree to which observed findings in a study lead to correct inferences about the source population.
Internal
____ validity refers to the degree to which study results can be generalized to the broader population.
external
What is the main question that internal validity addresses?
A) How well do the study results relate to the source population?
B) How well can the study results be generalized to the target population?
C) How accurately are sampling units defined?
D) How large should the sampling frame be?
A
What is the main question that external validity addresses?
A) How well do the study results relate to the source population?
B) How well can the study results be extrapolated to the target population?
C) How accurately were variables measured in the study?
D) How significant are the observed differences in the study?
B
A study has poor ____ validity if important confounding variables, such as smoking, are not accounted for when investigating the relationship between coffee drinking and cancer.
internal
External validity is more important than internal validity.
False (Internal validity is more important because results cannot be generalized if they are not valid in the first place).
A study with good external validity but poor internal validity is still reliable.
False
Confounding variables, such as smoking in a study about coffee drinking and cancer, can undermine ____ validity.
internal
What is the relationship between internal and external validity in epidemiologic research?
Internal validity is necessary for external validity. Without accurate results (internal validity), findings cannot be generalized to the broader population (external validity).
Non-probability sampling involves choosing study subjects in a ____ way, meaning not every member of the source population has an ____ probability of being selected
non-random, equal
____ sampling involves the investigator choosing units they deem representative of the population.
Judgment
Which form of sampling involves selecting units specifically due to their exposure or disease status?
A) Convenience sampling
B) Judgment sampling
C) Purposive sampling
D) Random sampling
C
Why is purposive sampling often used in analytical studies?
It ensures enough subjects with rare exposures or rare outcomes are included, which might not happen with other sampling methods.
(True/False): Convenience sampling is appropriate for populations that are homogenous.
True
Which of the following is NOT a benefit of non-probability sampling?
A) Easy to conduct
B) Cheap if subjects are chosen for convenience
C) Provides highly representative results for all populations
D) Useful for studying homogenous populations
C
____ sampling is a form of non-probability sampling where subjects are chosen because of ease of access.
convenience
Non-probability sampling is appropriate for populations that are very _______
homogenous
You are studying a rare disease and need enough participants who are already diagnosed. Which non-probability sampling method would you use?
Purposive sampling.
What are the five main types of random sampling?
Simple random sampling, systematic random sampling, stratified random sampling, cluster sampling, and multistage sampling.
(True/False): In simple random sampling, all individuals have an equal chance of being included in the study.
True
If the sampling interval is 5 and the first subject is randomly chosen, how are the remaining subjects selected in systematic random sampling?
Every 5th subject is selected after the first.
(True/False): Stratified random sampling involves dividing the sampling frame into sub-categories, or strata, before sampling.
True
Match the type of random sampling to its description.
Simple random sampling
Systematic random sampling
Stratified random sampling
A) Every individual has an equal chance of being selected.
B) Sampling involves dividing the population into sub-categories based on factors like age or gender.
C) A sampling interval is used to select every nth subject after the first random selection.
1-A, 2-C, 3-B
Cluster sampling uses ____ as the sampling unit, while the unit of concern is still the ____
groups (clusters), individual
In a study sampling herds, investigators select farms and examine all cows within the selected herds. What type of sampling is this?
Cluster sampling.
Multistage sampling differs from cluster sampling because it involves sampling at both the ____ level and the ____ level.
cluster, individual
Which of the following is NOT a pro of random sampling?
A) Provides a representative sample if done properly.
B) Easy to implement without identifying the sampling frame.
C) Useful for many different types of populations.
D) Robust and reliable when well-implemented.
B
To calculate the required sample size, we need three key components: estimated ____, level of ____, and
variance, confidence, precision
(True/False): Decreasing the allowable error (e.g., from 10% to 2%) will increase the required sample size.
True
What happens to the sample size if the desired precision increases (e.g., allowable error decreases)?
The sample size increases because a smaller allowable error requires a larger sample to achieve greater precision.
In analytical studies, the ____ hypothesis states no difference between groups, while the ____ hypothesis predicts a difference.
null, alternative
____ is the ability of a study to detect differences between groups when a real difference exists. It is typically set at ____%.
Power, 80
(True/False): A Type I error occurs when we incorrectly reject the null hypothesis when it is true.
True
(True/False): A Type II error occurs when we fail to reject the null hypothesis when it is false.
True
Match the error type to its description.
Type I error
Type II error
A) Falsely concluding there is a difference.
B) Failing to detect a difference when one exists.
1-A, 2-B
What happens to the sample size if the expected difference between two groups becomes larger?
The required sample size decreases because larger differences are easier to detect.
(True/False): It is easier to detect a small association between an outcome and an exposure variable than a large one.
False
What happens to the required sample size if the groups being studied are very similar in the variable of interest?
The required sample size increases because it is harder to distinguish differences between similar groups.
Required sample size ____ as the size of the difference between two means or proportions ____.
increases, decreases
