Overview
This lecture covers the essentials of survey research design, including question construction, types of survey questions, sampling methods, sources of bias, and best practices for psychological research.
Introduction to Survey Research
- Survey research is widely used in social sciences, especially psychology, for gathering information about people's attitudes, behaviors, and experiences.
- Surveys can dispel myths, reveal unexpected relationships, and inform policy or intervention decisions.
Designing and Using Questionnaires
- The primary goal in survey design is to clearly define what you hope to accomplish.
- Using existing questionnaires allows comparability, but unique topics may require designing your own instrument.
- The first step is to determine the specific information needed from the survey.
Types of Survey Questions
- Survey questions are open-ended (allow any response) or closed-ended (provide set choices).
- Open-ended questions provide richer data but are difficult to code and analyze on a large scale.
- Closed-ended questions are easier to analyze but may oversimplify complex issues.
Writing Good Survey Questions
- Each question should be clear, specific, and address only one concept (avoid double-barreled or ambiguous wording).
- Avoid leading or biased wording to ensure unbiased responses.
- Ensure closed-ended response options are mutually exclusive and exhaustive.
Common Biases and Response Effects
- Social desirability bias occurs when respondents answer in ways they think are socially acceptable.
- Verification keys (e.g., lie scales) can detect dishonest or biased responses.
- Acquiescence bias is the tendency to agree with statements, regardless of content, and affects binary (yes/no) questions.
Scaling and Survey Formats
- Likert scales and visual analog scales capture degrees of agreement or intensity.
- More scale points provide more information but may encourage neutral (midpoint) responses.
- Use branching items to make surveys efficient by skipping irrelevant questions for some respondents.
Survey Administration Methods
- Common methods: computerized (in-person or online), paper and pencil, and telephone.
- Each method has unique challenges (e.g., environment control, handwriting legibility, respondent authenticity).
Response Rates and Sampling
- Response rates vary by method: face-to-face surveys have the highest, magazine/online the lowest.
- Sampling methods: haphazard (non-systematic), purposive (selecting a specific group), convenience (easiest to reach), probability (random selection).
- Define your population and sampling frame clearly before data collection.
Probability Sampling Techniques
- Simple random sampling: every member has equal, independent chance of selection.
- Stratified random sampling: divides population into subgroups and samples proportionately from each.
- Cluster sampling: selects entire groups (e.g., classes, zip codes) and samples within them.
- Multistage sampling: combines cluster and random sampling in stages for efficiency.
Key Terms & Definitions
- Open-ended question — allows respondents to answer in their own words.
- Closed-ended question — provides predetermined response options.
- Mutually exclusive — categories that do not overlap.
- Exhaustive — categories that cover all possible responses.
- Social desirability bias — tendency to give socially acceptable answers.
- Acquiescence — tendency to agree with statements regardless of their content.
- Sampling frame — the specific list or group from which a sample is drawn.
- Simple random sample — every individual has an equal, independent chance of being selected.
- Stratified random sample — random sampling within identified subgroups.
- Cluster sample — random selection of groups, then sampling within those groups.
- Multistage sampling — combining multiple sampling methods across stages.
Action Items / Next Steps
- Respond to this week’s discussion questions posted on Canvas.
- Prepare for a deeper dive into statistical analysis in the next lecture.