Steps in the Data Analysis Process - Sampling
Sampling Technique

Simple Random

Conducted so that each person in the population has an equal chance of being selected

  • usually representative of the population
  • easy to analyze results
  • easy to understand 
  • requires enumeration of the population 
  • larger sampling error than stratified 

Teacher assigns each student a number.  She uses a random number generator to select 4 students from the class.

Stratified Random

Divides the population into subgroups and then randomly samples from within each group.

  • allows subgroup comparisons
  • fewer subjects needed
  • more representative of the population than other probability methods 
  • requires subgroup identification
  • requires knowledge of the proportional distribution of each subgroup 

To select a sample with the same racial makeup as the population, each patient is placed a category and 2% of the members of each group are picked at random


The population is divided into clusters; clusters are randomly selected and then the entire population of those clusters are selected for the sample.

  • each element in the cluster must be identified
  • efficient for large populations
  • difficult to get data from all elements of each cluster
  • requires that each element be assigned to only one cluster 

The entire congregations of ten churches in Iowa are sampled.

Systematic Sampling

Every kth member of the population is selected.

  • simplicity in drawing sample 
  • once the first subject is chosen, the rest of the subjects do not have an equal chance of being selected 

A cashier asks every 5th person in line if they want to contribute to a cancer fund.


Sample consists of subjects with easy access to the researcher

  • less costly 
  • less time-consuming
  • high participation rate
  • possible to generalize for similar subjects 
  • difficult to generalize to other subjects
  • less representative

A researcher asks students in his psychology class which soda they prefer.


subjects are selected so that various subgroups are represented without knowing the exact probability of being selected

  • all pros of convenience sampling
  • most representative of non-probability sampling 
  • all cons of convenience sampling 
  • most time-consuming of non-probability sampling 

To get a variety of perspectives on religious tolerance, the researcher selects 10 catholics, 10 muslims and 10 atheists. 


subjects are intentionally selected on the basis of the wealth of information they can provide to the researcher 

  • all pros of convenience
  • grantees gathering of needed information 
  • all cons of convenience sampling 

Researcher selects 20 cancer patients for whom a treatment has been successful. 


Back to the Steps in the Data Analysis Process