Why We Conduct Polling
At the Colorado Polling Institute (CPI), our scientific public opinion polling research is intended to provide reliable, data-driven insights for the public, the media, decision-makers, and elected officials on the views of Coloradans.
Well-conducted polling helps ground debates in evidence rather than anecdotes, ensuring that policy discussions are informed by what voters actually think and experience.
Why These Standards Matter
Polling is not about predicting outcomes or declaring winners—it is about measuring public opinion as accurately as possible at a moment in time. Using scientific methods helps minimize error, reduce bias, and ensure that all segments of the population have a voice.
CPI’s work aligns with the professional and ethical standards established by organizations such as the American Association for Public Opinion Research (AAPOR), which sets best practices for survey transparency, methodology, and research ethics.
What Makes a Poll “Scientific?”
Not all polls are created equal. CPI adheres to widely accepted best practices in survey research to ensure our results are accurate, credible, and useful. This requires intentionality & investment.
Key characteristics of scientific polling include:
Random Sampling
The cornerstone of scientific polling is random sampling. In a random sample, every individual in the target population—in most CPI surveys, registered or likely Colorado voters—has a chance of being selected to participate.
To maximize participation and reduce bias, CPI uses multiple modes of data collection, such as phone, text-to-web, and online methods, allowing respondents to engage in the way that is most accessible to them.
Representativeness
Even with random sampling, some groups are more likely to respond to surveys than others. To address this, CPI uses statistical adjustments known as weighting.
Weighting aligns the final sample with the known demographic profile of the electorate—such as age, gender, education, race and ethnicity, geography, and party registration—based on high-quality public data. This process helps ensure that the survey reflects the broader population, not just those most inclined to respond.
Clear and Unbiased Questionnaire Design
Scientific polling depends on carefully written questions. CPI surveys use clear, neutral, and balanced wording to avoid leading respondents or introducing bias. Questions are designed to be easily understood and to measure opinion—not persuade or influence it.
When possible, we also test alternative wordings and question orders to ensure results are not driven by survey design choices.
Bipartisan Research Team
To further guard against bias, CPI’s polling is conducted by two professional research firms from opposite sides of the partisan spectrum. This bipartisan structure helps ensure that survey questions, response options, and interpretations are scrutinized from multiple perspectives and designed to fairly measure public opinion.
Transparency
Transparency is central to CPI’s mission. For every poll we release, we publicly disclose detailed methodological information, including:
- Dates of interviewing.
- The population that was sampled (e.g. registered voters or likely voters).
- Sample size and the margin of error.
- Complete question wording and response options.
- Mode(s) used to conduct he interviews (e.g., phone, text, online).
This allows journalists, researchers, and members of the public to independently evaluate the quality and limitations of our work.