Chapter 10: Population and Sampling

Sampling

n Process of selecting representative units (a portion) of a population for research

n PURPOSE: to increase the efficiency of a research study

Population

n Well-defined set that has specific properties

n Can be people, animals, objects, or events

– patient records

– specimens: blood, urine

– historical documents

Eligibility

n Eligibility criteria are used to select the sample from the population

n Define characteristics that should be included or excluded

Sample

n Set of elements that make up the population

n Element: the most basic unit about which information is collected

n Representative sample

– One whose key characteristics closely approximate those of the population

Types of Sampling

n Probability or Random

n Nonprobability

– Convenience

– Quota

– Purposive

Probability sampling

n Uses some form of random selection when choosing the sample units

n Each element of population has an equal and independent chance of being chosen

n Bias eliminated

n Inconvenient, not easy

Probability Sampling:
Random Sampling

Four types of random selection

n Simple random sampling

n Stratified random sampling

n Cluster sampling

n Systematic sampling

Nonprobability sampling

n Nonrandom methods, less representative

n No way of insuring that every element has an equal chance of being included Less rigorous, easier to get, bias

n Tends to produce less accurate and less representative sample

n Volunteers, convenient

n Limits ability to generalize

Nonprobability Sampling: Convenience Sampling

n Most readily accessible persons or objects

n Risk of bias because group is self-selecting (what motivates them)

n Volunteers (likely to differ from population being studied - not representative)

Nonprobability Sampling: Quota Sampling

n Knowledge of information about population is used to build some representativeness into the sample

n Avoids over or under representation of subjects

n Assign subjects by proportion in the population

Nonprobability Sampling: Purposive Sampling

n Knowledge of population elements is used to hand-pick subjects to be included in study

n Selection of those who are typical of population

n Limits ability to generalize

Sample Size

n Larger is better (for quantitative research)

– More representative

n 30 subjects for every variable

n Power analysis to determine size

Randomization

n Random sampling -selection of a sample from the population

n Random assignment- Unbiased assigning of subjects from a sample into groups in an experimental study

n Experimental treatments are randomly assigned to groups

THE END