National Methodology

Morning Consult fields the general population polls online, among a national sample of adults (age 18+) living in the United States (including the District of Columbia) using non-probability sampling. For every survey, Morning Consult draws a stratified sample from Purespectrum and other online sample exchanges, based on age by gender quota targets derived from the U.S. Census Bureau’s 2024 Current Population Survey.

Both sample exchanges integrate with a variety of sample providers and implement their own quality control and deduplication methods (in addition to Morning Consult’s own quality control checks) for all providers to ensure unduplicated supply of pre-screened respondents. Sample exchanges act as a marketplace where research companies’ demand and sample providers’ supply are matched via bidding, and Morning Consult enters bids for the desired sample per group with a pre-selected (white-listed) number of sample providers. For EdChoice’s monthly tracking poll, approximately 2,200 adults are interviewed in English each month with a field period of 2-5 days.

In addition to the monthly general population polling, Morning Consult also conducts a monthly online oversample poll of K-12 school parents, among a national sample of parents living in the United States (including the District of Columbia). The sample is collected via stratified sampling from PureSpectrum and other online sample exchanges, based on race and ethnicity quota targets derived from the 2023 American Community Survey. Approximately 800 K-12 school parents are interviewed in English over one week each month. The completed general population and K-12 school parent interviews are weighted using iterative proportional fitting (raking) without trimming to population totals obtained from the 2023 American Community Survey on the following demographic variables: age, gender, region, race/ethnicity, and educational attainment.

In addition to the monthly polling, Morning Consult also conducts a separate, semiannual online poll of teachers, among a national sample of teachers living in the United States (including the District of Columbia). The sample is collected via stratified sampling from PureSpectrum and other online sample exchanges, based on gender by years of teaching experience quota targets derived from the U.S Department of Education, National Center for Education Statistics, 2020-2021 National Teacher and Principal Survey. Approximately 1,000 teachers are interviewed in English over one week twice a year. The completed teacher interviews are weighted using iterative proportional fitting (raking) without trimming to population totals obtained from the 2020-2021 National Teacher and Principal Survey on the following demographic variables: age, gender, race/ethnicity, and teaching experience.

State-level respondent data are pooled on a monthly rolling basis in order to report state data and results. Once 12 months of data are collected, the pooled data will consist of the most recent wave/month plus the prior 11 months of respondent data – producing a 12-month rolling average. We do not report state results if the cumulative monthly sample totals less than 50 respondents. And we advise caution on interpreting the polling results for any state or population sample size having less than 100 respondents.

For quality control purposes, Morning Consult includes a set of attentiveness tests including simple, open-ended questions and a semantic association question. Respondents who complete the survey too quickly or show suspicious response patterns are removed from the results before weighting and do not count towards the total sample size.

Full survey specifications, such as dates, median survey times, sample sizes, measures of precision, and completion rates for individual polls can be found in the monthly/quarterly reports located on the Resource Downloads page. The Completion Rate is reported as the number of panel members who completed the survey divided by the number of panel members who were invited to take the survey, based on AAPOR outcome rate guidelines for non-probability internet panels.

Morning Consult reports an adjusted measure of sampling error to estimate the precision of the results. This calculation (like all other uncertainty measures for non-probability samples) assumes that the weighted estimates are approximately unbiased. This approximate unbiasedness assumption is based on the assertion that any systematic differences between sample and population are corrected when the sample is adjusted via weighting to match the population on key dimensions (age, race/ethnicity, education, gender, region for the general population, and gender and years of teaching experience for the teacher sample). While Morning Consult regularly validates its weighting methodology on benchmarks derived from U.S. Census data, no additional analysis is done to validate the assertion for the tracking polls.

The reported uncertainty for these polls was calculated using the classical formula for sampling error and inflated the estimate using the design effect of the study to incorporate the effect of weighting, defined as:

This measure reflects the uncertainty of reported results due to sampling and weighting under the assumption of approximate unbiasedness: Using the same methodology and absent any systematic selection effects, 95 out of 100 times the reported proportions of interest will fall within this estimated uncertainty range. The uncertainty of the reported estimates at the proportion of interest of 50% for these polls was approximately 2.38 percentage points and will be larger for subgroups.

EdChoice is committed to research that adheres to high scientific standards, and matters of methodology and transparency are taken seriously at all levels of our organization. We are dedicated to providing high-quality information in a transparent and efficient manner.

The American Association for Public Opinion Research (AAPOR) welcomed EdChoice to its AAPOR Transparency Initiative (TI) in September of 2015. The TI is designed to acknowledge those organizations that pledge to practice transparency in their reporting of survey-based research findings and abide by AAPOR’s disclosure standards as stated in the Code of Professional Ethics and Practices.