1.4 Types of Data: Primary vs Secondary

A research team must decide between collecting new measurements themselves or using an existing dataset. Which choice most effectively balances timeliness and awareness of limitations against cost and effort?

• Use outdated public reports to save more time

• Combine secondary summaries without checking meanings

• Rely on secondary data compiled by another source

• Use primary data collected with a tailored method

A university team wants to study lecturers’ motivation and distributes questionnaires to a sampled list of lecturers they compiled. Which data type are they collecting, and why?

• Primary data, derived from historical archives and reports

• Secondary data, gathered by interviewing third-party analysts

• Primary data, collected directly from sampled respondents

• Secondary data, extracted from existing institutional records

An engineer has a homogenous population of 900 sensors listed in a sampling frame. She needs a sample of 30 using Simple Random Sampling. Which approach best ensures each sensor has an equal chance of selection?

• Select sensors from one factory line

• Choose first 30 sensors on list

• Pick every 30th sensor from list

• Draw 30 labels from a sealed box

A factory has a population of N = 1,200 units and needs a sample of n = 60 using systematic random sampling. Which first step correctly sets up the sampling interval k and why?

• Set k = 60 because the sample size equals interval

• Compute k = n/N = 0.05 to space selections evenly

• Compute k = N/n = 20 to space selections evenly

• Choose k arbitrarily between 1 and 60 for flexibility

A city has 12,000 residents divided into three strata: 7,200 engineers, 3,600 technicians, and 1,200 managers. You plan a stratified random sample of 240 people using proportional allocation. How many technicians should be included?

• 80 technicians selected proportionally

• 72 technicians selected proportionally

• 60 technicians selected proportionally

• 48 technicians selected proportionally

Which statement best describes the design principle of stratified random sampling when forming strata?

• Strata and elements both uniformly heterogeneous everywhere

• Strata internally heterogeneous, population across strata homogeneous

• Strata internally homogeneous, population across strata heterogeneous

• Strata randomly formed without considering characteristics

A city has 40 residential areas, each with 50 households. To select a sample of 200 households efficiently, which design best matches cluster sampling and why?

• Divide areas into strata and sample some households from each

• Randomly pick 200 households across all areas proportionally

• Select 4 areas randomly and survey all households in those areas

• Systematically choose every 10th household across the entire city

A citywide housing survey will use multistage sampling. First, a subset of states is randomly chosen; within those, districts are randomly selected; finally, housing estates are sampled. Which statement best explains why this approach is appropriate for very large populations, as depicted in the diagram?

• It reduces selection bias by eliminating subclusters entirely

• It allows progressive narrowing to manageable sampling units

• It guarantees homogeneous groups at every stage automatically

• It increases precision by measuring all units within clusters

A market researcher must survey 200 adults with quotas matching city demographics but can select any accessible individuals to fill each subgroup. Which approach best minimizes sampling bias in this scenario?

• Snowball sampling from initial volunteer participants

• Strict random selection within each subgroup quotas

• Judgmental selection within quotas near the researcher

• Convenience sampling of shoppers in one mall

A factory wants to monitor defect rates. It selects every 50th item off the conveyor after a random start. Which sampling technique is being used, and does it fall under random or non-random sampling?

• Systematic sampling, non-random category

• Systematic sampling, random category

• Cluster sampling, random category

• Quota sampling, non-random category