Americans are sharply divided on standardized testing and many other hot button educational issues, two new back-to-school public opinion polls find. But which way they lean seems to depend on who asks the question and how they phrase it.
Testing lacks public support, reads one headline in the annual Phi Delta Kappan/Gallup poll report released this week. PDK/Gallup found that 64 percent of public school parents think there is too much emphasis on standardized testing in public schools.
Public Backs Testing, reads the headline in a competing poll released the previous week, this one sponsored by the Harvard-based education reform group Education Next. The Ed Next poll found that 66 percent of parents favor keeping the federal requirements that children be tested annually in math and reading in grades 3 through 8 and once in high school.
Experts agree both polls are likely accurate, but answers hinge on how questions are phrased and advocates on either side will often spin those answers to say more or less than they actually do.
The answers you get depend heavily on the way the question is posed, acknowledged Harvard education professor Martin West, a co-author of the Education Next report, on a conference call.
"You can get any result you want," agrees Bob Schaeffer, public education director for the Boston-based advocacy group FairTest, which opposes heavy use of standardized tests. "If you asked whether every student should be tested annually with results used to grade schools and teachers, you would get a very different answer, he said.
In fact, poll questions often are specifically designed not to reveal what the public thinks, but to uncover what it might think if offered specific cues. A reader who looks for nuances in the phrasing of such polls, and does not take anything for granted, can thus learn a lot about both what Americans thinks and the strategies advocates might employ to nudge them.
Nudges
West says that the Education Next poll intentionally alternates some questions to find what Americans think about new curriculum standards.
For instance, for one question, they used the brand name Common Core. The alternate question is phrased exactly the same, except that it refers to standards for reading and math that are the same across the states, rather than Common Core. The goal was to see if the public had become hostile to the Common Core brand or to the concept of nationwide curriculum standards.
The words made some difference in this case, with 47 percent of parents favoring Common Core and 41 percent opposing. Meanwhile, 49 percent favored the generic policy without the trigger phrase, and 34 percent opposed it.
But the PDK/Gallup poll headline finds the Common Core out of favor, with 54 percent of public school parents opposed and 25 percent favoring.
Why the difference?
In a phone conversation, Paul Peterson, one of the coauthors of the Education Next report and a Harvard public policy professor, thinks the answer lies in the question wording.
In fact, the wording in the PDK poll is quite neutral. It asks if they would require teachers to use the Common Core state standards to guide what they teach.
The Ed Next poll, on the other hand, nudges the listener by suggesting that states use the Common Core to hold public schools accountable for their performance. Its an inflection that matters: no one wants unaccountable schools.
School choice
Another area where the two polls split is charter schools. But in this case, question phrasing is not the culprit. The PDK poll found that 64 percent favored "the idea of charter schools." But only 48 percent in the Ed Next poll favored "the formation of charter schools."
Peterson and West in a blog post explain part of this discrepancy by looking in a place where few would think to look. They note that 35 percent chose "neither" in the Ed Next poll, but just 11 percent chose "don't know" in the PDK poll, and surmise that people are reluctant to admit ignorance.
It is worth noting that the Ed Next poll uses a five point scale that allows for leaners and neutrals, where the PDK poll offers only two choices.
On the other key piece of school choice, tuition vouchers, the polls lined up nicely.
Do you favor or oppose allowing students and parents to choose a private school to attend at public expense? asked the PDK poll. It found 57 percent opposed.
The Ed Next poll uses a large sample to test four different question phrasings on vouchers. Of the four, the one that drew the most negative response was the one identical to the PDK phrasing, which emphasizes public cost (over parental choice) and universal vouchers (over low-income targeted vouchers). It showed 58 percent opposed.
The most popular Ed Next voucher question emphasized parental choice over public cost and offered vouchers to all parents. That question had 46 percent supporting and 36 percent opposed.
Testing troubles
On the other hand, Peterson objects to a PDK question on allowing parents to opt their children out of required tests. The question asked if parents should be allowed to excuse their children from standardized exams.
Peterson sees a similar nudge in the PDK questions on standardized tests. Is there too much emphasis on standardized questions? the poll asks.
On both questions, the word standardized, Peterson argues, has negative associations. And in the second question, the phrase too much also primes the listener with negative emotions.
We excuse kids to go to the bathroom or go home when they are sick, Peterson said. To excuse something sounds reasonable and offers an air of legitimacy to the question.
On three questions using the standardized phrase, the PDK poll found parents leery of testing. PDK reports that 63 percent believe there is too much emphasis on standardized exams, 55 percent opposed using them to evaluate teachers, and just 44 percent opposed allowing parents to opt their children out of them.
In contrast, the Ed Next poll found that 66 percent of parents favored federal requirements for annual testing in math and reading in grades 3-8, and 52 percent of parents opposed allowing parents to opt their children out.
Looking for clues
It's useful to note that these polls don't fall silently in a forest. There are really people battling over real educational issues who are watching them very closely.
On the testing opt out issue, for example, a large national movement has taken root with parents and teachers passionately committed to what amounts to civil disobedience in some cases.
Bob Schaeffer points to a Sienna College poll in New York state that found that 50 percent favored allowing parents to opt out, against 44 percent opposed, quite different results from the national poll. Schaefer is not surprised at the discrepancy, noting that New York citizens may also be more informed about the issue. Estimates are that 200,000 children opted out of state tests in New York this year.
And down in Florida, Cindy Hamilton is watching the polls with a wary eye, looking for clues in the phrasing that can guide her strategy. Hamilton is the co-founder of Opt Out Orlando, a flash point in Florida's bookend to New York's opt out movement this year.
Hamilton recognizes that the significance of the polls is not what they say about a fixed creature called public opinion, but rather what information they offer people on the field. She takes comfort, she said, in the notion that the polling questions on her issue did not capture its nuances.
"Our challenge is to convince people," Hamilton said.
Testing lacks public support, reads one headline in the annual Phi Delta Kappan/Gallup poll report released this week. PDK/Gallup found that 64 percent of public school parents think there is too much emphasis on standardized testing in public schools.
Public Backs Testing, reads the headline in a competing poll released the previous week, this one sponsored by the Harvard-based education reform group Education Next. The Ed Next poll found that 66 percent of parents favor keeping the federal requirements that children be tested annually in math and reading in grades 3 through 8 and once in high school.
Experts agree both polls are likely accurate, but answers hinge on how questions are phrased and advocates on either side will often spin those answers to say more or less than they actually do.
The answers you get depend heavily on the way the question is posed, acknowledged Harvard education professor Martin West, a co-author of the Education Next report, on a conference call.
"You can get any result you want," agrees Bob Schaeffer, public education director for the Boston-based advocacy group FairTest, which opposes heavy use of standardized tests. "If you asked whether every student should be tested annually with results used to grade schools and teachers, you would get a very different answer, he said.
In fact, poll questions often are specifically designed not to reveal what the public thinks, but to uncover what it might think if offered specific cues. A reader who looks for nuances in the phrasing of such polls, and does not take anything for granted, can thus learn a lot about both what Americans thinks and the strategies advocates might employ to nudge them.
Nudges
West says that the Education Next poll intentionally alternates some questions to find what Americans think about new curriculum standards.
For instance, for one question, they used the brand name Common Core. The alternate question is phrased exactly the same, except that it refers to standards for reading and math that are the same across the states, rather than Common Core. The goal was to see if the public had become hostile to the Common Core brand or to the concept of nationwide curriculum standards.
The words made some difference in this case, with 47 percent of parents favoring Common Core and 41 percent opposing. Meanwhile, 49 percent favored the generic policy without the trigger phrase, and 34 percent opposed it.
But the PDK/Gallup poll headline finds the Common Core out of favor, with 54 percent of public school parents opposed and 25 percent favoring.
Why the difference?
In a phone conversation, Paul Peterson, one of the coauthors of the Education Next report and a Harvard public policy professor, thinks the answer lies in the question wording.
In fact, the wording in the PDK poll is quite neutral. It asks if they would require teachers to use the Common Core state standards to guide what they teach.
The Ed Next poll, on the other hand, nudges the listener by suggesting that states use the Common Core to hold public schools accountable for their performance. Its an inflection that matters: no one wants unaccountable schools.
School choice
Another area where the two polls split is charter schools. But in this case, question phrasing is not the culprit. The PDK poll found that 64 percent favored "the idea of charter schools." But only 48 percent in the Ed Next poll favored "the formation of charter schools."
Peterson and West in a blog post explain part of this discrepancy by looking in a place where few would think to look. They note that 35 percent chose "neither" in the Ed Next poll, but just 11 percent chose "don't know" in the PDK poll, and surmise that people are reluctant to admit ignorance.
It is worth noting that the Ed Next poll uses a five point scale that allows for leaners and neutrals, where the PDK poll offers only two choices.
On the other key piece of school choice, tuition vouchers, the polls lined up nicely.
Do you favor or oppose allowing students and parents to choose a private school to attend at public expense? asked the PDK poll. It found 57 percent opposed.
The Ed Next poll uses a large sample to test four different question phrasings on vouchers. Of the four, the one that drew the most negative response was the one identical to the PDK phrasing, which emphasizes public cost (over parental choice) and universal vouchers (over low-income targeted vouchers). It showed 58 percent opposed.
The most popular Ed Next voucher question emphasized parental choice over public cost and offered vouchers to all parents. That question had 46 percent supporting and 36 percent opposed.
Testing troubles
On the other hand, Peterson objects to a PDK question on allowing parents to opt their children out of required tests. The question asked if parents should be allowed to excuse their children from standardized exams.
Peterson sees a similar nudge in the PDK questions on standardized tests. Is there too much emphasis on standardized questions? the poll asks.
On both questions, the word standardized, Peterson argues, has negative associations. And in the second question, the phrase too much also primes the listener with negative emotions.
We excuse kids to go to the bathroom or go home when they are sick, Peterson said. To excuse something sounds reasonable and offers an air of legitimacy to the question.
On three questions using the standardized phrase, the PDK poll found parents leery of testing. PDK reports that 63 percent believe there is too much emphasis on standardized exams, 55 percent opposed using them to evaluate teachers, and just 44 percent opposed allowing parents to opt their children out of them.
In contrast, the Ed Next poll found that 66 percent of parents favored federal requirements for annual testing in math and reading in grades 3-8, and 52 percent of parents opposed allowing parents to opt their children out.
Looking for clues
It's useful to note that these polls don't fall silently in a forest. There are really people battling over real educational issues who are watching them very closely.
On the testing opt out issue, for example, a large national movement has taken root with parents and teachers passionately committed to what amounts to civil disobedience in some cases.
Bob Schaeffer points to a Sienna College poll in New York state that found that 50 percent favored allowing parents to opt out, against 44 percent opposed, quite different results from the national poll. Schaefer is not surprised at the discrepancy, noting that New York citizens may also be more informed about the issue. Estimates are that 200,000 children opted out of state tests in New York this year.
And down in Florida, Cindy Hamilton is watching the polls with a wary eye, looking for clues in the phrasing that can guide her strategy. Hamilton is the co-founder of Opt Out Orlando, a flash point in Florida's bookend to New York's opt out movement this year.
Hamilton recognizes that the significance of the polls is not what they say about a fixed creature called public opinion, but rather what information they offer people on the field. She takes comfort, she said, in the notion that the polling questions on her issue did not capture its nuances.
"Our challenge is to convince people," Hamilton said.
