Chapter 7 Survey Research - Ask people to provide information - Important complement to experimental research Closed-ended question - Respondents get limited number of response options - Converting the response options to numerical data that can easily be analyzed - Easier to code - Used when variables are well-known / defined - Might miss key information Rating scale - Normal 5 or 7 point likert scale - Graphic rating: mark across a continuous line with two opposite views - Semantic differential scale: 7-point scale with two opposing views in context of evaluation - Most commonly used: evaluation e.g. good vs bad, high vs low, kind vs cruel, wise vs foolish - Second: activity e.g. active vs passive, slow vs fast, excitable vs calm - Third: Potency e.g. weak vs strong, hard vs soft, large vs small - Non Verbal scale, images instead of word/ numbers. perhaps for children who don't understand yet - High-frequency scale vs low-frequency scale: most of the options measure a behaviour done very frequently vs not as frequently Open-ended question: - Harder to code - Richer information - Used in qualitative research (data based) - Used when exploring new topic The exact same question can be open-ended or closed-ended Questionable wording (Avoid) Double-barreled question: two or more conflict/ contradiction element Loaded questions: Participants feel be judged when answering the question Negative wording: more than one negation e.g. not disapprove Yea-saying/Naysaying Inattentive responding, providing answers without reading the item content Use reversed questions/Conscientious responders scale to avoid Response Bias Response set: a pattern of responding to questions that is not related to the content of the questions themselves and thus provides inaccurate information Social desirability: most likely to arise when questions concern a sensitive topic (race, political) To avoid: ask questions in a way that hides the main intent. e.g. How often do you physically hit others? How often do you encounter people who are just asking to be hit? Ensure privacy Fence-sitting: be neutral for everyone Remove the neutral option to avoid Done using special programs or instruments. e.g. smart phone apps, EAR Administrating surveys in interview format People are more likely to answer questions Give interviews chances to clarify questions But, interviewer bias: summarizes all of the biases that can arise from the fact that the interviewer is a person interacting with another person.
Face to face interviews: - Used when sample size is small, and benefits are obvious - Expensive and time-consuming Telephone/ online interview - Fast, results are easy to analyze Focus group interview: A focus group is an interview with a group about 6-10 people brought together for a period of usually two to three hours. - People can respond to one another, may elicit more talking - More Open-ended questions - Need to be recorded and transcribe - Time-consuming and costly - Provide tons of information Population: the group of people of interest to the researchers, from which a sample is typically drawn (sampled) Sampling: the process of choosing members of