176 15.5 Survey Sample Types: Non-Probability

You must be very careful to evaluate the sampling techniques employed in any survey you are contemplating using. There are many types of samples, and some are more reliable than others. The two types of survey samples are non-probability samples and probability samples.

NON-PROBABILITY SAMPLES

Non-probability sampling methods include all those in which respondents are selected without randomness. Non-probability sampling is generally cheaper and easier to do than probability sampling. However, a survey that uses a non-probability sample generally has no value to a communicator interested in revealing information about a particular population.

Types of Non-Probability Samples

Available or “convenience” sample

• Interviewers choose respondents who are readily accessible

  • Using people passing by a street corner

• There is no way to determine who the respondents represent

• Results of the survey are not generalizable to any larger group

  • a convenience sample conducted at a mall in an upscale neighborhood will not yield results that can be represented as the opinions of “typical” consumers for a particular product

Volunteer sample

• Interviewees choose to volunteer for the sample

  • Using students who volunteer to complete a survey

  • Posting a questionnaire on a popular website or social media platform and inviting responses

  • Using a “call-in” survey on a TV or radio show

• People who volunteer for things usually have different characteristics from those of the general population (more education, higher intelligence, greater need for approval). This makes them unrepresentative of a larger group

• Impossible to generalize from these results to any larger group

Quota sample

• Interviewers choose respondents based on prearranged categories of sample characteristics

  • Setting out to interview 50 men and 50 women for a survey and considering the survey completed when those numbers have been reached

• Does not allow the pollster to measure the representativeness of any larger group

Purposive sample

• Interviewers select respondents from subgroups in the population in question because these subgroups have specific characteristics or qualities of interest

  • At a grocery store interviewing only those people who say they eat yogurt and not administering the survey to any others

• Results can reveal the opinions of a specific subset, but cannot be representative of any larger group