Research Design & Methodology Flashcards

A researcher wants to establish that a new training program caused improved employee performance, not other factors. Which design best supports this causal claim?

Answer: Randomized controlled experiment

Randomized controlled experiments allow causal inference by controlling for confounders through random assignment.

In a longitudinal panel study, attrition bias occurs when:

Answer: Dropouts differ systematically from those who remain

Attrition bias arises when participants who leave the study differ from those who stay, skewing results.

Which sampling method is most appropriate when a complete population list is unavailable and the population is geographically dispersed?

Answer: Cluster sampling

Cluster sampling divides the population into groups and randomly selects entire clusters, ideal when a full list is unavailable.

A researcher uses the same instrument to measure customer satisfaction today and again in six months without any intervention. This assesses:

Answer: Test-retest reliability

Test-retest reliability measures the consistency of an instrument's results over time without any intervening changes.

What is the primary purpose of a pilot study?

Answer: To test and refine instruments and procedures before the main study

A pilot study identifies flaws in instruments, procedures, and logistics before the full study is conducted.

Convergent validity is demonstrated when:

Answer: A measure correlates highly with measures of theoretically related constructs

Convergent validity is established when a measure correlates strongly with other measures of the same or closely related constructs.

In experimental design, counterbalancing is used to control for:

Answer: Order effects and carryover effects

Counterbalancing varies the order of conditions across participants to cancel out order and carryover effects.