Research Center

Three takeaways from this blog about our referendum predictions:

1. They are accurate about 94% of the time.

2. They have been, on average, within 0.7% of the final election outcome.

3. They are accurate because of the way we write.

Not long after I was hired, a school district sent me the following text to include in their opening letter. (I altered the names and dates to protect the not-so-innocent.)

The Lake View School District strives to exemplify astute fiscal responsibility and judicious management of taxpayer funds. Since the revenue limit institution during the 1993 state legislative session, we have consistently levied amounts below this threshold, thereby circumventing the tertiary aid penalties embedded within the general state aid formula. During the previous referendum, we employed prudent financial forecasts when estimating the cost of constructing modern, innovative learning environments for our scholars. Consequently, the final project costs were below the anticipated budget. With the emergence of favorable interest rates in the late 2010s, the Board implemented an accelerated debt repayment strategy through defeasance. Coupled with robust property value growth, the mill rate impact from the 2010 referendum is presently 31% lower than initially projected. This financial prudence implies that Lake View could issue $21.32 million in bonds subject to long-term, tax-exempt interest rates due to our AAA MMD rating while maintaining a mill rate below the original projection.

An online reading level calculator told me this passage is “difficult to read, best understood by university graduates, and is appropriate for an academic paper.”

Umm, yeah. I would agree.

At the time, though, I didn’t have a data-driven way of explaining to them why this was a bad idea. It is. But I didn’t have the words. (Probably because they took them all.)

Now, we do.

With that in mind, I want to answer two questions for you:

1. How accurate is School Perceptions when making a referendum prediction from a survey?

2. How does writing the way we do make our accuracy rate so high?

What is our accuracy rate?

Since 2018 and across nearly 400 referendum survey projects, our referendum predictions have been accurate approximately 94% of the time. That doesn’t mean your referendum is going to pass. Instead, it means what the data says will happen almost always happens.

Put differently, if the results tell us a referendum will pass, and it does, that’s a win. If the results tell us a referendum will fail, a district takes that plan to ballot anyhow, and it fails, that’s a win too.

However, that doesn’t tell the whole story. It doesn’t tell you how right we are.

Think of it this way: let’s say we predict a referendum will pass with 71% support, but on election day, the referendum passes with 51% support. We predicted pass, and it did. We were right. But how you communicate about a potential landslide referendum versus how you communicate about a potential nailbiter referendum will be different. You need to know by how much a potential referendum will pass or fail.

Sometimes, our predictions are a little too high, and sometimes, they’re a little too low. However, on average, since 2020, our referendum predictions have been within 0.7% of the ballot outcome.

How does the way we write contribute to our accuracy?

Survey writing is not beautiful writing. Save that for literature. Survey writing “clearly conveys specific information to busy people and makes it easy for them to understand and respond.”*

We aim to write every survey using words all residents, especially Grandma and Grandpa, can understand. If respondents understand the problem the district is facing and the possible solutions, you will have good data to make decisions.

We keep the following ideas in mind and what they mean for survey development.

The research	What it means for you
Our brain decides within 17 milliseconds if we like what we clicked on.	The first few words of your survey need to be relatable to the average person, and they need to get to the point ASAP.
The average person reads 265 words per minute and checks their phone 344 times per day – once every four minutes.	You get 1,060 words before someone checks their phone, and we all know apps are designed to distract us and suck us in as long as possible. How many vital points can you make before someone finds something more interesting to do?
Writers who use simpler words actually come across as more intelligent than those who use fancier words.	You’re not going to impress anyone with big or jargon-y words. Don’t bother. Readers respond to authenticity and relatability. Be simple, straightforward, and conversational. Would you hear people saying what you’re typing at the 6:00am coffee clutch or a 6:00pm fish fry? If not, delete it. Rewrite it. And read what you say out loud. People talk more clearly than they write.
Only half of employees reported reading notes from their own bosses.	So why should they listen to you? Tell them something they don’t know, need to know, or should know. Then, imagine your survey respondent asking, “Why me? Why did I receive this message?” Then, put those words first in your survey. Use the US Army’s principle: bottom line up front.
Writers are selfish.	Writers write with themselves in mind. We think about what we want to say, not what others would and should hear. When everything is important, nothing is important. Every time you revise, ask yourself, “What is the most important information I want my readers to understand?” and “How do I make it easier for my readers to understand it?”
You’ll gain more respect when you show respect.	People get very annoyed if you waste their time. Give readers the option to learn more on your website or a separate FAQ. This option shows that you respect them, you respect their time, you’re on their side, and you want to make it as easy as possible for them to learn what they need to know.
Just because people care doesn’t mean you’ll keep their attention.	Even your biggest supporters will ignore what you have to say if you have a) too much text, b) too much jargon, or c) too many choices. If you lose a reader, it’s not their fault.
About 50% of Americans read at an eighth-grade level or below.	Just because a reader has less education doesn’t mean they’re any less deserving to know what’s happening at school. They are members of your community, they pay taxes, and they vote.

If you’re considering a community survey this fall, reach out. The sooner, the better!

Credit for the research and concepts above goes to:

Rogers, T. & Lasky-Fink, J. (2023). Writing for Busy Readers: Communicate More Effectively in the Real World. Dutton.

VandeHei, J., Allen, M., & Schwartz, R. (2022). Smart Brevity: The Power of Saying More with Less. Workman Publishing Company. 

---

The School Perceptions Blog and Resource Center features the voices of our team members. This post was written by Rob DeMeuse, Research Director.