COMM 101 Lecture Notes - Lecture 14: American National Election Studies, Simple Random Sample, Convenience Sampling

• School: U of M
• Department: Communication Studies
• Course: COMM 101
• Professor: Julia Lippman

Document Summary

Survey: a questionnaire designed to measure the variables related to a speci c set of hypotheses (or research questions) How do you conduct a survey: ask questions to assess the concepts you want to measure, get respondents to answer, compare responses to variables to evaluate hypotheses. Take away #1: not all surveys are scienti c. Types of survey mode: in-person surveys (face-to-face): interviewer interacts with respondent in-person, can be done with non-probability or probability samples. Probability: american national election studies (cluster sampling) Good: response quality is usually very good, very low non-response, respondents are willing to do a longer interview. Non-probability: call people from a directory (convenience sampling) Probability: random digit dialing/rdd (simple random sampling from phone numbers. Good: generally good response quality, low cost per respondent for probability sample, surveys can be conducted quickly. Probability: address-based sampling (simple random sampling from postal addresses)