Over the past 18 months, Bellwether’s analysis of parent polling data shows that families want different educational experiences for their kids. The sector must support parent preferences, which requires engaging in new and thoughtful analyses that paint a more complete picture of parents’ priorities.
Across the country, parents want to see significant changes in K-12 education post-pandemic. That’s not conjecture or wishful thinking: it’s borne out by a large, longitudinal collection of data on parents’ preferences, enrollment trends, and focus groups. The data reveal key insights for policymakers, advocates, and journalists:
- Parents want access to more educational options, but are hindered by financial, informational, and other barriers.
- There is significant unmet demand among parents for a range of supplemental and extracurricular opportunities, but similar barriers constrain participation.
- Policymakers in states of all political makeups can make substantive progress to reduce barriers and increase access to a range of educational options that parents want for their children.
These takeaways are a good starting point for the education sector, but more analysis is needed to develop policies and build the options families want moving forward. Much like the insights above, the answers won’t be found in the toplines of a single poll — they’ll come through thoughtful data collection and rigorous, ongoing analysis.
Bellwether learned important lessons about how to become good consumers of data on parent preferences:
- Interpret (or frame) questions carefully: How a question is asked matters. The framing of questions and response options can push responses in one direction or another. Check to see if a poll’s framing is neutral in tone or if it might trigger a more partisan or otherwise biased response.
- Analyze data in context: Individual poll results may be useful, but it’s important to understand how they fit within the context of other data on the topic. When analyzing poll data, look for other data on similar questions across different polls and over time.
- Invest in building a deeper understanding through qualitative data: There’s only so much that can be gleaned from a fixed set of multiple-choice questions. Open-ended responses and focus groups are an important complement to traditional public opinion polling that can provide more nuanced and unexpected insights.
- Be wary of psychological biases: Asking about sensitive or emotional topics can be influenced by biases like cognitive dissonance and self-presentation. This is an important dynamic to consider when surveying parents about their children’s education.
- Compare stated to revealed preferences: Actions speak louder than words — including those found in survey results. Comparing parents’ stated preferences (via surveys) to their revealed preferences (via their actions) provides a deeper understanding of how, or if, polling data reflects actual behavior.
- Use opinion research to generate new questions: Polling can provide important data points, but it can’t provide perfect clarity on the path forward. Instead, combining polling with focus groups and direct parent outreach efforts can help our sector ask better questions of parents so that solutions more closely reflect their needs.
These practices are useful in any context and are increasingly important in K-12 education as rising polarization distorts perceptions of public sentiment. Below, we describe our approach in greater detail to serve as a resource for advocates, thought leaders, and journalists as we all work to empower and support families in getting their children the high-quality education they deserve.
Read on for more about what we learned.
A Closer Look:
After several years of disrupted schooling amid the COVID-19 pandemic, the depth and diversity of educational needs facing families and students are greater and more complex than ever. The stakes for building the educational recovery solutions that families need couldn’t be higher or more urgent.
Education policymakers must confront a series of thorny and interrelated problems: massive learning loss, significant public school enrollment declines, and looming fiscal challenges. The consequences for not addressing these challenges, and learning loss in particular, will be massive: The COVID-era cohort of students is expected to lose more than 5% of its lifetime earnings.
A one-size-fits-all path forward won’t cut it for families. Instead, they need access to a range of options customized to meet each child’s needs. To make that possible, the news media, advocates, and policymakers must develop a deeper understanding of what we know (and what we don’t know) about parents’ preferences.
Polling is a useful, though imperfect, source of insight. Discerning what parents want from schools and other education providers is challenging; whether you’re an advocate, policymaker, or journalist, finding time to directly engage with a broad cross-section of parents is daunting, if not impossible. Our sector often turns to public opinion polls to get a representative sense of how parents feel about a range of educational issues because it’s the most obvious source of such data. But interpreting that data can be more challenging than it seems; insights are valuable but often incomplete.
Minimizing statistical bias in poll data is a challenge that’s bedeviled professional pollsters for decades. Polling firms deploy a range of methods — from shifting from phone to online formats and weighting different demographic categories to achieve representative samples — but as we’ve seen in the last few election cycles, it’s still an imperfect science.
Over the past 18 months, Bellwether analyzed information on what parents want from the country’s K-12 education system in the wake of the COVID-19 pandemic. Of course, there’s much more to understand about parents’ priorities. Our sector must examine efforts to elevate parents’ voices through thoughtful polling, analysis, and engagement in six ways.
Interpret (or frame) questions carefully
The way poll results spread through headlines and social media feeds often obscures the nuance in how questions and responses are presented to survey respondents. For instance, questions that are posed through a partisan lens may reveal more about a respondent’s underlying political leanings than the question itself. Political polarization is playing a larger role in how people think about all facets of life, including education issues, and more people are starting to respond to polling questions from the perspective of their political affiliation or changing their education views to fit their partisan preferences. This is part of why the “partisan gap” in public opinion polling is growing across a range of education issues from spending to charter schools to teachers unions.
Take care to look at how poll questions and response options are phrased: Is the language neutral in tone or could it skew responses in one direction? Using loaded words (see page 84) can prime survey participants to respond in a particular way and may say more about how they respond to terminology than what they think on a particular issue. Sector leaders can elevate the debate on important issues by focusing on poll results from questions that avoid phrasing that is easily translated to nationalized political topics. PIE Network (in partnership with Echelon Insights) and More in Common have shown how this can be done to understand opinions about race and U.S. history instruction without using questions with culture war phraseology.
Analyze data in context
The way survey items are structured within the broader context of a poll can affect how its results are interpreted. If a survey asks a series of questions that address hot-button issues followed by otherwise neutral phrasing on education issues, it may prime respondents to interpret well-worded items through a particular lens. For example, it’s not hard to imagine that parents may express more concern about their child’s academic progress after answering a series of questions about school closures and learning loss during the pandemic.
Polling data should also be considered in the context of other polls that have asked similar questions. A single poll may produce an interesting result in its toplines or cross tabulations, but it’s best to consider how it fits in the broader context of data on similar questions over time. Looking at other polls that address the same issue and comparing their wording and results can help confirm a trend where there is convergence with other data or identify an outlier when there is divergence.
The risk of overinterpreting a single data point at a single moment in time is why Bellwether developed the Parent Perception Barometer, an interactive tool that aggregates data on parents’ priorities to add clarity and context to conversations about what parents think about K-12 education. The Barometer’s parent polling data separates signal from noise by allowing users to see longitudinal trends that emerge across multiple polls.
Polls provide a moment-in-time picture of parent preferences; the context of those moments should also be considered when analyzing poll data. For instance, polling of parent opinion in the weeks following the school shooting in Uvalde, Texas indicated heightened parent concerns regarding a violent intruder in their child’s school that declined over time.
Invest in building a deeper understanding through qualitative data
There’s a broad recognition in the education sector that while standardized, multiple-choice exams can tell us something, different data sources are needed to paint a more complete picture of K-12 student learning. The same goes for parent opinion — to get beyond the constraints of Likert-scale responses, different methods add depth and understanding.
Qualitative research provides an opportunity to understand the deeper motivations behind parents’ stated preferences, which can inform future quantitative research. For example, a focus group of parents who were frustrated by the education available to their children during the first year of the pandemic expressed concerns about making a high-stakes decision to switch schools but were interested in incremental change. These insights guided our research on extracurricular opportunities, highlighting the types of activities that attract the most parent interest and the degree to which logistics, cost, and lack of information limit participation.
Be wary of psychological biases
Polls are often influenced by psychological biases when they delve into sensitive or emotionally charged topics. This is a particular challenge when trying to learn about parents’ views about their children’s education.
Even an innocuous question about the quality of education their children receive can illustrate these challenges among a sample set of parent respondents. Self-presentation bias may cause more parents to overstate their confidence in the quality of education their child is receiving, because they don’t want to admit that they aren’t providing the best possible learning opportunities. Cognitive dissonance — discomfort with new information that challenges previously held beliefs — may also influence how parents respond to such a question, even if they had misgivings about the quality of their child’s education. There is also evidence that social desirability bias may influence how people respond to pollsters on sensitive issues, regardless of whether or not they hold those opinions.
People are reluctant to admit when they make a mistake purchasing a phone plan or an appliance — imagine how fraught it is to second-guess high-stakes decisions about their child’s education.
Compare stated to revealed preferences
Polling data can present conflicting information about parents’ views on K-12 education issues. From 2005 to 2022, Gallup’s tracking of parents’ satisfaction with their oldest child’s education held steady around 76% with very little deviation — even during the pandemic. On the surface, this statistic might lead you to believe that most parents are satisfied with the status quo in schooling. But other data complicates that narrative. Several polls from 2020 to 2022 consistently show that more than half of parents want to see “bold change” in schools or “rethink” how schools educate students.
One way to clarify these stated preferences is to examine revealed preferences. In other words, how do survey data reflect — or contrast — with parents’ actual behavior? Parents may be more willing to sound off to a pollster on the phone about their frustrations with their child’s school than to move their child to a different school.
Parents are acting on their actual preferences. Public school enrollments are down more than one million students since the start of the pandemic as families sought new schooling options at rates much higher than pre-pandemic levels. Charter schools, private schools, home-schools, and other innovative models enabled by a slew of policies like education savings accounts have experienced increased enrollment post-pandemic. Parents are voting with their feet and it’s vital that the K-12 sector understand and respond to their shifting preferences. This data on revealed preferences (e.g., shifts in enrollment or program participation) provide deeper context on how parent’s stated preferences align with their willingness to act.
Use opinion research to generate new questions
Polling is a helpful, but not definitive tool. It’s more of a topographical map than your phone’s GPS: polling can reveal the contours of the landscape, but it won’t necessarily direct the “right” course of action. News media, advocates, and policymakers should avoid the temptation to champion or rely on polling statistics as absolute truths. Instead, polling, focus groups, and direct parent outreach efforts should inform future lines of inquiry to pursue to better understand shifting parent preferences.
Consider the apparent disconnect between data on COVID-era learning loss and parents’ concerns about their child’s academics. On one hand, data from the National Assessment of Educational Progress provides insight into the scale and scope of learning loss: significant drops in reading and math performance, expanding gaps based on income and race, and an increase in students performing at the lowest performance category (Below Basic). Data from other test vendors indicate similar trends. On the other hand, parent polling data from 2020 to 2022 indicates a decrease in parent concern about academics as schools returned to in-person instruction. In cases like this, polling provides answers to some questions (e.g., “how concerned are parents about their child’s academic progress?”) but raises new ones that future polls or other research could help answer (e.g., “what information do parents use to understand their child’s academic progress?”).
Similarly, Bellwether’s analyses on parent opinion answer some questions and raise others. However, poll data, qualitative insights, and enrollment and other data on parent behavior give us confidence in two things: (1) millions of parents want more than a return to the pre-pandemic status quo for their kids, and (2) millions of parents are interested in more incremental ways to customize their children’s education.
Those insights raise areas of additional inquiry into the complexity of the diverse interests and priorities parents have for their children’s education. What does “better” mean to parents? What incremental changes are parents interested in, and what prevents them from making wholesale shifts? What prevents parents from accessing what they need and want for their kids? How can sector leaders address those barriers? There are no easy or monolithic answers to these questions. Deeper research into parents’ opinions can help journalists, advocates, and policymakers understand and respond to the contours of parents’ concerns.
As the education sector pursues thoughtful engagement to understand trends in parents’ priorities, particularly post-pandemic, we should move beyond the gravitational pull of individual or partisan data points to truly understand what parents want and the solutions their children need. The Parent Perception Barometer will be an ongoing source of reliable information to do just that.