Rendition (1)

Sample Surveys

  • Sample surveys involve collecting data from a subset of a population to make inferences about the entire population.

  • Errors can lead to bias in survey results.

Types of Errors in Sample Surveys

  • Two main types of errors: sampling errors and non-sampling errors.

Sampling Errors

  • Defined as errors caused by the act of taking a sample.

  • A sampling error results from the deviation of the selected sample from the true characteristics of the population.

  • Caused by:

    • Random sampling causes minor errors due to chance.

    • Bad sampling methods (e.g., voluntary response sampling, convenience sampling).

Random Sampling Error
  • Results from chance deviations between the sample statistic and the population parameter.

  • Can be controlled by increasing the sample size.

Non-Sampling Errors

  • Errors unrelated to selecting a sample; more difficult to manage and quantify.

Frame Errors
  • Occur when the sampling frame does not accurately represent the population.

  • Example: Using telephone directories for surveys leads to undercoverage.

  • Undercoverage excludes certain groups:

    • Those with unlisted landline numbers.

    • Individuals without landline phones (e.g., low-income households, young people).

    • Those only having cell phones.

  • To mitigate, use random digit dialing to achieve a better sample.

Nonresponse Errors
  • The failure to obtain data from selected individuals for a sample.

  • Considered the most serious problem in sample surveys.

  • Managed through substitution:

    • Replace nonrespondent households with other households from the same neighborhood.

Response Errors
  • Occurs when respondents provide incorrect answers, often due to a desire to appear favorable.

  • Example: Individuals lying about alcohol consumption or smoking habits.

  • To reduce, ask sensitive questions carefully, ensuring confidentiality.

Processing Errors
  • Mistakes in mechanical tasks, such as calculations or data entry.

  • The use of computers has reduced the frequency of these errors.


Sample Surveys: Collect data from a subset to infer about the entire population. Errors can lead to bias.

Types of Errors:

  1. Sampling Errors: Caused by the act of taking a sample; deviations from true population characteristics.

    • Caused by random chance and poor sampling methods (e.g., convenience sampling).

    • Random Sampling Error: Can be mitigated by increasing sample size.

  2. Non-Sampling Errors: Not related to sample selection; harder to manage.

    • Frame Errors: Sampling frame misrepresents the population (e.g., using phone directories).

    • Undercoverage: Excludes certain groups; can be improved with random digit dialing.

    • Nonresponse Errors: Missing data from selected individuals; most serious problem. Substitute when possible.

    • Response Errors: Incorrect answers due to social desirability. Mitigate through careful questioning.

    • Processing Errors: Mistakes in calculations/data entry; less common due to computer use.