FAST Sampling Methods and Design 3 — Questions and Answers
Question 1: A sample statistic is used to estimate a population parameter. As sample size increases, what generally happens to the sampling error?
- It increases proportionally
- It remains constant
- It decreases (Correct answer)
- It becomes impossible to calculate
Correct answer: It decreases
Larger samples reduce sampling error because they provide more information and more closely approximate the population.
Question 2: In experimental design, a 'control group' serves what primary purpose?
- To receive the experimental treatment at a higher dose
- To provide a baseline for comparing the effect of the treatment (Correct answer)
- To increase the total sample size of the study
- To eliminate the need for random assignment
Correct answer: To provide a baseline for comparing the effect of the treatment
A control group does not receive the treatment, establishing a baseline against which treatment effects can be measured.
Question 3: A study randomly assigns participants to treatment and control groups. What is the main advantage of random assignment?
- It guarantees a larger sample size
- It eliminates all measurement error
- It helps balance known and unknown confounding variables across groups (Correct answer)
- It ensures participants complete the study
Correct answer: It helps balance known and unknown confounding variables across groups
Random assignment distributes both known and unknown confounders roughly equally between groups, supporting causal inference.
Question 4: Which term describes a variable that is related to both the independent and dependent variables and may distort the apparent relationship between them?
- Mediator
- Confounder (Correct answer)
- Moderator
- Covariate suppressor
Correct answer: Confounder
A confounder is a third variable that can create a spurious or masked association between the independent and dependent variables.
Question 5: A convenience sample drawn from one university cannot be generalized to all college students in the U.S. This limitation is a problem of:
- Internal validity
- Construct validity
- External validity (Correct answer)
- Statistical conclusion validity
Correct answer: External validity
External validity concerns whether findings can be generalized beyond the specific sample, setting, or time of the study.
Question 6: A researcher divides a city into blocks, randomly selects 20 blocks, and then interviews all residents within those blocks. This is best described as:
- Stratified random sampling
- Simple random sampling
- Single-stage cluster sampling (Correct answer)
- Multistage cluster sampling
Correct answer: Single-stage cluster sampling
Single-stage cluster sampling randomly selects clusters and then surveys all members within those clusters.
Question 7: A study's results are said to have high internal validity when:
- The sample closely mirrors the broader population
- The measurement tool is consistent across administrations
- Changes in the outcome can be attributed to the independent variable (Correct answer)
- The sample size exceeds 1,000 participants
Correct answer: Changes in the outcome can be attributed to the independent variable
Internal validity refers to the degree to which a study establishes a causal relationship between the independent variable and the outcome.
FAST Sampling Methods and Design 3
A sample statistic is used to estimate a population parameter.
As sample size increases, what generally happens to the sampling error?