• the purpose of the survey is to produce statistics— quantitative descriptions about some aspects of the study population

• the main way of collecting info is by asking people questions

• info is generally collected about a fraction of the population (a sample)

Reasons for surveys

• three potential properties of data from a properly done survey that may make them preferable to data from other sources

  • probability sampling

    • enables one to have confidence that the sample isn’t a biased one and to estimate how precise the data are likely to be

  • standardized measurement

    • consistent across all respondents

    • ensures that comparable info is obtained about everyone who’s described

  • special-purpose survey

    • ensures that all the data for a given analysis are available and can be related

Components of surveys

• sampling

  • to select a small subset of a population representative of the whole population

  • the key is finding a way to give nearly all population members the same chance of being selected and using probability methods for choosing the sample

• question design

  • researchers evaluate questions to find out if they are well understood and if the answers are meaningful

  • pretests of surveys

  • choice of question wording- more objective

• interviewing

  • when interviewers are used, it’s important that they avoid influencing the answers respondents give and at the same time maximize the accuracy with which questions are answered

  • give the interviewers standardized questions

  • interviewers need to be trained in how to administer a survey to avoid introducing important biases in the answers they obtained

• mode of data collection

  • in-person

  • telephone interviewing

  • the internet

  • mail surveys

• total survey design

  • optimal survey design will take into account all the most important aspects of the survey process

    • whether or not to use a probability sample

    • the sample frame (people who have a chance to be sampled)

    • the size of the sample

    • the sample design

    • the response rate

  • the quality of the data will be no better than the most error-prone feature of the survey design