POLS 1602 02.17

Poll Types in Political Surveying

Trend Polls

  • Definition: A trend poll typically surveys fewer than 800 people and lessens the number of questions asked.

  • Purpose: Used as follow-ups to baseline polls, assessing issues that resonate with the population.

Tracking Polls

  • Definition: Tracking polls capture evolving public opinion over time, acting like "moving pictures" rather than static snapshots.

  • Sampling Process:

    • Initially sample 200 people per night.

    • On each subsequent night, 200 new individuals are sampled, while the earliest responses are dropped.

    • This gives a changing sample size, approximately totaling 600 individuals at any one time.

  • Importance: Helps candidates gauge public reaction to their activities such as advertisements and events.

    • Initiated a few days prior to major campaign actions to analyze public awareness and reactions.

Exit Polls

  • Definition: Conducted on Election Day to predict electoral outcomes based on voter exit responses.

  • Warning: Polls offer estimates with a margin of error; they should not be mistaken for actual election results.

  • Sample Selection:

    • Pollsters choose precincts believed to be representative of the broader electoral landscape.

    • Diverse demographic representation is sought, including age, gender, and ethnicity to provide a comprehensive analysis of voting trends.

    • Pollsters also account for voting times to ensure varied demographics are surveyed (morning, midday, evening).

Focus Groups

  • Definition: Unlike polls, focus groups consist of non-random, targeted demographics (20 usually) intended to provide qualitative insights rather than quantitative data.

  • Composition:

    • Often segmented by specific demographics (e.g., age, gender, occupation) to ensure comfort and candid feedback from participants.

  • Methodology:

    • Participants engage with campaign messaging, advertisements, and discuss impressions openly.

    • Real-time feedback is gathered through methods like dialing scales (right for positive, left for negative), visually tracking reactions.

  • Importance: Enables campaigns to refine messaging and gauge audience perceptions dynamically.

  • Data Analysis: Information is categorized by demographic segments, allowing tailored adjustments to campaign strategies.

Push Polls

  • Definition: Push polls masquerade as surveys but aim to influence or alter public opinion rather than measure it.

  • Characteristics: Often start with benign questions and shift to leading, biased queries designed to sway opinions.

  • Example: They present fictitious narratives about candidates, highlighting negative traits in opponents while exaggerating positive traits.

Challenges and Limitations of Polls

  • Number of Issues Raised:

    • Response Set: When pre-defined responses influence actual sentiment.

    • Order Effects: The sequence of questions can impact how respondents answer.

    • Temporal Limitations: Opinions can shift rapidly in response to current events.

    • Low Response Rates: Small response pools can skew findings and may not represent the overall population.

    • Margin of Error: Understanding estimates (+/- 3% for around 1,100 respondents) and how they affect interpretation of results.

  • Confidence Intervals: Overlap in candidate approval ratings indicates no significant statistical difference; a critical aspect for election interpretations.

Application: The Husky Poll Results

  • Conducted standard Gallup question: Sample indicated 24% satisfied, 77% dissatisfied with U.S. state.

  • Class population dissatisfaction markedly higher than U.S. population: 85% dissatisfied, illustrating potential regional perceptions.

  • Approval Ratings for Political Figures: Discussion on approval ratings and comparisons between respondents’ demographics within the classroom versus wider citizenry.

Presenting Data in Political Surveys

  • Pie Charts, Bar Graphs, and Line Graphs: Understanding the most appropriate visualization tools for different types of data representation.

    • Pie Charts: Useful for categorical data;

    • Bar Graphs: Better for ordinal data (e.g., approval levels);

    • Line Graphs: Least appropriate for non-time-series data.

Political Identification Measurement

  • Questions designed to gauge partisan identification and individual sentiment towards political affiliations.

    • Breakdown of responses from different demographic groups (Democratic, Republican, Independent, and their respective approval levels).

    • Discussion around the prominence of party identification affecting voter behavior, including responses based on party alignment.

Conclusion

  • Heavy emphasis on the significance of sampling methods, question constructions, and timing of polls for accurate public opinion measurement.

  • Critical need for awareness regarding potential biases and data limitations in political polling.

  • Reflective exercises on polling participation: insights into how polling influences understanding of broader political trends and individual engagement in politics.