upgrade
upgrade

🪛Intro to Political Research

Survey Question Types

Study smarter with Fiveable

Get study guides, practice questions, and cheatsheets for all your subjects. Join 500,000+ students with a 96% pass rate.

Get Started

Why This Matters

Survey design sits at the heart of political research—it's how we measure public opinion, test hypotheses, and gather the data that drives political analysis. You're being tested not just on recognizing different question types, but on understanding when and why researchers choose one format over another. The choice between a Likert scale and a ranking question isn't arbitrary; it reflects fundamental decisions about measurement validity, data analysis, and the kind of conclusions you can draw.

Think of question types as tools in a methodological toolkit. Each one captures different dimensions of opinion—intensity, preference, presence/absence, or open expression—and each comes with trade-offs between depth and standardization, nuance and comparability. When you encounter survey methodology on exams, don't just identify the question type—ask yourself what the researcher gains and loses by using it, and what kind of analysis it enables.


Closed-Ended Questions: Standardization and Comparability

Closed-ended questions provide respondents with predefined answer options, making responses easy to code, compare, and analyze statistically. The trade-off is clear: you gain reliability and comparability at the cost of respondent freedom and potential insight into unexpected perspectives.

Multiple Choice Questions

  • Predefined response options—respondents select from a fixed set of answers, either single-answer or multiple-answer format
  • Quantitative analysis ready—responses translate directly into frequencies and percentages for statistical comparison
  • Reduces measurement error—eliminates ambiguity in interpretation compared to open-ended formats, though may force artificial choices

Dichotomous Questions (Yes/No)

  • Binary response structure—offers exactly two options, typically yes/no or agree/disagree
  • Screening function—commonly used as filter questions to route respondents to relevant follow-up items
  • Limited depth—sacrifices nuance for clarity; useful for establishing presence/absence of characteristics but not intensity

Closed-Ended Questions (General Category)

  • Umbrella category—encompasses multiple choice, Likert scales, dichotomous, and other fixed-response formats
  • Statistical analysis—enables quantitative research through easy coding and aggregation across respondents
  • Validity trade-off—reduces misinterpretation risk but may miss perspectives not anticipated by researchers

Compare: Multiple Choice vs. Dichotomous—both are closed-ended, but multiple choice allows gradation among options while dichotomous forces a binary split. Use dichotomous for screening; use multiple choice when you need to distinguish among several discrete categories.


Scaled Questions: Measuring Intensity and Degree

Scaled questions move beyond simple categories to capture how much respondents agree, prefer, or feel about something. These formats recognize that opinions exist on a continuum and provide the ordinal or interval data needed for more sophisticated analysis.

Likert Scale Questions

  • Intensity measurement—captures degree of agreement/disagreement, typically on 5- or 7-point scales from "strongly agree" to "strongly disagree"
  • Attitudinal research standard—the most common format in social science for measuring opinions, beliefs, and perceptions
  • Cross-group comparison—enables researchers to compare attitude intensity across demographics, time periods, or experimental conditions

Rating Scale Questions

  • Numeric evaluation—asks respondents to score items on a defined scale (e.g., 1-10 satisfaction rating)
  • Flexible application—adaptable to measure satisfaction, frequency, importance, or quality depending on research needs
  • Quantitative output—produces numerical data suitable for calculating means, comparing groups, and tracking changes over time

Semantic Differential Scale Questions

  • Bipolar adjective pairs—measures attitudes using opposing terms (good-bad, strong-weak, active-passive) on a continuum
  • Connotative meaning—captures the feeling or association respondents have with concepts, not just cognitive evaluation
  • Brand and image research—particularly valuable in political communication for assessing candidate image or policy framing effects

Compare: Likert Scale vs. Semantic Differential—both measure intensity, but Likert focuses on agreement with statements while semantic differential captures associative meaning through adjective pairs. If an FRQ asks about measuring candidate image, semantic differential is your strongest example.


Preference and Priority Questions: Revealing Trade-offs

Some research questions require understanding not just what people think, but what they prioritize when forced to choose. These formats reveal the relative importance respondents assign to competing options—essential for understanding political decision-making.

Ranking Questions

  • Forced prioritization—requires respondents to order items from most to least preferred or important
  • Reveals trade-offs—shows what respondents sacrifice when they can't have everything, unlike rating scales where everything can score high
  • Ordinal data complexity—produces rankings rather than scores, requiring specialized analysis techniques like rank-order correlation

Matrix Questions

  • Grid format efficiency—presents multiple related items with identical response options in a single visual array
  • Dimensional measurement—useful for assessing attitudes across several related aspects (e.g., rating multiple policy areas on importance)
  • Response fatigue risk—overuse leads to satisficing behavior where respondents select the same answer repeatedly without careful consideration

Compare: Ranking vs. Rating Scale—rating scales let respondents evaluate each item independently (all could be "very important"), while ranking forces differentiation. Use ranking when you need to understand priorities; use rating when you want absolute evaluations. This distinction frequently appears in methodology questions.


Open-Ended and Adaptive Questions: Depth and Flexibility

Not all research questions can be anticipated in advance, and not all respondents fit neatly into predefined categories. Open-ended and contingency questions sacrifice standardization for richer, more tailored data collection.

Open-Ended Questions

  • Qualitative depth—respondents answer in their own words, revealing perspectives researchers may not have anticipated
  • Exploratory research value—ideal for pilot studies, understanding complex issues, or capturing the language respondents naturally use
  • Analysis burden—requires coding and interpretation, making large-scale quantitative analysis time-consuming and potentially subjective

Contingency Questions

  • Conditional follow-up—questions that appear only based on responses to previous items (e.g., "If yes, please explain...")
  • Targeted depth—allows drilling into specific subgroups without burdening irrelevant respondents with inapplicable questions
  • Survey logic design—requires careful skip-pattern planning to maintain flow and ensure all respondents receive appropriate items

Compare: Open-Ended vs. Closed-Ended—this is the fundamental trade-off in survey design. Open-ended provides depth and discovery; closed-ended provides comparability and efficiency. Strong research often uses both: closed-ended for hypothesis testing, open-ended for exploration and illustration.


Quick Reference Table

ConceptBest Examples
Measuring attitude intensityLikert Scale, Rating Scale, Semantic Differential
Binary/categorical classificationDichotomous, Multiple Choice
Revealing priorities and trade-offsRanking Questions
Efficient multi-item measurementMatrix Questions
Qualitative depth and explorationOpen-Ended Questions
Adaptive survey flowContingency Questions
Standardized quantitative analysisAll Closed-Ended formats
Connotative/emotional meaningSemantic Differential

Self-Check Questions

  1. A researcher wants to understand which policy issues voters consider most important when they can only address three. Which question type best reveals these priorities, and why wouldn't a rating scale work as well?

  2. Compare Likert scale and semantic differential questions: both measure attitudes, but what different dimensions of attitude does each capture?

  3. You're designing a survey about candidate image. Which question type would best capture the feelings voters associate with a candidate (trustworthy-untrustworthy, strong-weak), and how does this differ from measuring policy agreement?

  4. A pilot study aims to discover what concerns voters have about a new policy—concerns the researchers may not have anticipated. Which question type is most appropriate, and what trade-off does this choice involve?

  5. Explain how contingency questions improve both respondent experience and data quality. Give an example of when a researcher would use one in a political survey.