Psychology 2020 Introduction to Psychological Methods Unit 3 Surveys!

Survey Research Surveys provide a methodology for finding out what people think at a particular time They provide a source of data that can complement data from experimental research It is assumed that people are able and willing answer survey questions truthfully and accurately

Survey Response Errors A response set is when a participant answers all questions on a survey from a particular perspective rather than honestly. –Social Desirability (faking good) Answering the questions in such a way as to make their answers like “most people” –“Yea-saying” or “Nay-saying” Answering all questions as either yes/agree or no/disagree regardless of the content. Researchers can reduce response set biases by being open and honest with the participants, provide feedback on the results and assure confidentiality.

Constructing Survey Questions Survey questions should be based on the research objective (what is it that the researchers want to know) and usually include questions about –Attitudes and beliefs –Facts and demographics –Specific behaviors (past & future) Question wording needs to be easy to understand and neutral (non-leading)

Wording for Questions Avoid questions that ask for two things at once (double-barreled) –“Should students pay less for tuition and have fewer classes offered each semester?” Avoid loaded questions –“ Do you support the useless recall election for Governor of California?” Avoid negative wording –“Parents who don’t abuse their children don’t have low self-esteem” Change wording to reduce “yea and nay-saying”

Responses to Questions Closed- versus open-ended questions –Closed-ended questions are more structured and easier to score but can be missleading. –Open-ended questions can produce insightful information not predicted by the researcher Closed-ended questions include a variety of methods for quantifying response alternatives.

Questionnaires Advantages –Are less costly to administer than interviews in terms of researcher time. –Allow the responses to be completely anonymous. Disadvantages –Respondents must be able to read and understand the questions. –Could be boring and respondents won’t finish them or be motivated to answer the questions completely.

Interviews Advantages –Higher response rates because people are more likely to answer questions from a real person as opposed to a mailed questionnaire. –Rapport with interviewer motivates more complete answers and answering all the questions. –Interviewer can clarify the questions and ask follow-up questions. Disadvantages –Interviewer bias (expectations, approval/disapproval)

Sampling A population is all individuals of interest to a researcher –For identical twins research the population would be all individuals with an identical twin. A sample is a subset of the population that is selected to participate in the research. –200 identical twins randomly selected from the population participated in the research.

Confidence Intervals Data collected from the sample are used to make inferences about the variables of interest in the population. Inferences can contain error (sampling error) which is estimated and used to provide a confidence interval –A confidence interval is a range of values drawn from the sample that the researcher is 95% sure are true for the population. A larger sample size (up to a point) provides a more accurate prediction of the true population value.

Unbiased Sampling To insure an accurate sample, researchers must: –Randomly sample from a population that contains all members of that population –Collect data from ALL participants in the selected sample. Completely unbiased samples are rare and truly random samples are often difficult to obtain. –Researchers tolerate nonrandom samples because they are not trying to estimate population values but are investigating general relationships between variables within a population.

Sampling Techiques Probability Sampling –Each member of the population has a specifiable probability of being selected Simple random sampling: each person has an equal probability of being selected. Stratified random sampling: the total population is divided into relevant “strata” (or subdivisions) and then members of each strata are randomly selected Cluster sampling: the population is organized in clusters and then a subset of clusters is randomly selected, then all individuals within the cluster are sampled.

Sampling Techinques Nonprobability Sampling –No systematic method is used to insure the sample is representative of the population Haphazard sampling- select people wherever you can find them Quota sampling-pick representative percentages of people to represent the numerical composition of the population. –This is different from stratified and cluster sampling because there is no random selection but only haphazard selection.

Sampling Methods Terms Sampling Frame –The actual population of individuals or clusters from which the random sample is drawn. –It is important to consider how well the sampling frame matches the population of interest. Response Rate –The percentage of participants in the sample who completed the survey. –The lower the response rate the more likely the sample results will be biased.