A new Quinnipiac poll is out today. As always, I think it’s best to take these polls not as single data points in favor of one particular position, but rather as part of a broad sea of often contradictory, incoherent evidence about what/whether the public thinks about education.

There are some interesting nuggets in here, and again fodder for both “sides” of current education reform debates. The teachers’ unions and their supporters will love that the poll finds voters support the teachers’ unions’ policies over Governor Cuomo’s by a substantial margin (note that Cuomo’s overall favorability rating is net positive, so the lack of support for his education policies is particularly strong; that said, I wonder how much people understand what his education policies even are). The reformsters will love that a majority thinks the number of charter schools in the state should be expanded. Nothing new here; support for charters in polls is almost always net positive.

What’s most interesting to me, though is a series of questions about standardized testing. To me, these questions make painfully apparent the utter lack of coherence (or, to put it much more charitably, the nuance) in the public’s views of testing. First, we have the question “Do you think teacher pay should or should not be based on how well their students perform on standardized tests?” The results here are a resounding NO, 69/28. Similar results for whether standardized tests should be used for teacher tenure. [1]

Then we have the question “How much should these tests count in a teacher evaluation: 100%, 75%, 50%, 25%, or not at all?” Now, you would imagine these results would be mostly “not at all,” since the very previous two questions folks said the results shouldn’t be used for pay or tenure. Nope! In fact, 49% of people say these tests should count 50% or more in teacher evaluation, and another 27% say 25%. Just one-fifth of respondents–21%–say not at all. Hardly an anti-test bunch, these voters.

And finally, we have the question “Do you think standardized tests are or are not an accurate way to measure how well students are learning?” At this point I guess you’d have to think that voters would say yes, since in the immediately preceding question 77% said these tests should count for teacher evaluation. But you’d be wrong again! 64% said that standardized tests were NOT an accurate way to measure how well students are learning.

So, tests are not an accurate way to measure student learning, but they should definitely count at least a quarter in teacher evaluations, but they shouldn’t count at all in tenure or pay decisions. Got it. Suffice it to say this is yet another example showing why it’s immensely problematic when people pick a single data point from one poll and use it in support of their existing position.

[1] Jacob Mishook, on Twitter, notes that these wordings could be construed to imply 100% reliance on standardized tests for these decisions, which is a fair point that might explain at least part of the very negative response.