MARKET2

Population

the total group of interest to the researcher

Defined target population

the complete set of elements to be studied

Sampling unit

an individual element available for selection

Sampling frame

a list of all eligible sampling units

Sampling

selecting a subset of a population to represent the whole

Census

includes every member of the population

Why do we use sampling?

because studying the entire population is impractical or impossible

Sampling error

error due to the sample selection process

Nonsampling error

error not related to sampling (measurement, response, etc.)

Probability sampling

each unit has a known chance of being selected

Nonprobability sampling

selection probability is unknown

Simple random sampling

every unit has an equal chance

Systematic sampling

select every nth element (using a skip interval)

Stratified sampling

divide population into subgroups and sample from each

Cluster sampling

divide population into clusters and select some clusters

Convenience sampling

based on ease of access

Judgment sampling

based on researcher’s decision

Quota sampling

ensures representation of specific characteristics

Snowball sampling

participants refer others

Central Limit Theorem

large sample means → approximately normal distribution

Measurement

assigning meaning or numbers to concepts/objects

Construct

a hypothetical concept created in the researcher’s mind

Scale measurement

assigning descriptors to represent responses

Scale points

degrees of intensity in responses

Nominal scale

categories only, no order / gender, marital status

Ordinal scale

ranking, no equal distance / education level, preference ranking

Interval scale

equal distance, no true zero / temperature

Ratio scale

income, age, number of children

interval vs ratio

ratio has a true zero, interval does not

Semantic differential

bipolar scale (e.g., good–bad)

Behavioral intention scale

likelihood of future behavior

Noncomparative scale

evaluates one object independently

Comparative scale

compares objects

Rank-order scale

ranks preferences

Constant-sum scale

distributes points (e.g., 100 total)

Questionnaire

a formal set of questions designed to collect primary data

Structured question

closed-ended, fixed response options

Unstructured question

open-ended, respondent answers freely

Response order bias

answer influenced by order of options

Common method variance (CMV)

bias caused by measurement method