While you write the questions
Survey question readability checker
Check a survey question for reading level, sentence length, double-barrelled asks, negations, and vague quantifiers. Runs in your browser with no account.
No login required. Calculations run in your browser.
Reading level
Hard
Undergraduate level. Consider simpler wording.
- Flesch reading ease
- 36
- Grade level
- 13.4
- Words
- 21
- Words a sentence
- 21
Long words here: satisfied, friendliness, recently, customer, representatives.
4 things to look at
21 words a sentence
This sentence is long. Consider splitting it or moving background information into a separate instruction.
Possibly two questions in one
This may ask respondents to rate two things with one answer. If someone could give different answers to each part, split the question.
Vague quantifier: recently
Respondents may interpret this frequency or time term differently. Define it, or ask for a count within a stated period, such as the last 30 days.
5 long words
Long words are what push both scores. Most have a shorter twin that means the same thing to more of your sample: utilise/use, prior to/before, sufficient/enough.
How this tool works
Two scores, both standard. Flesch Reading Ease is 206.835 − 1.015 × (words ÷ sentences) − 84.6 × (syllables ÷ words); 90 and above is very easy, 60 to 70 plain, 30 and below very hard. The Flesch–Kincaid grade level is 0.39 × (words ÷ sentences) + 11.8 × (syllables ÷ words) − 15.59, and reads as a US school grade. Both are driven by two things only: how long your sentences are and how many syllables your words have.
Syllables are estimated from vowel groups, with adjustments for common English endings. Individual words can be miscounted, which can affect short questions noticeably. Sentences are counted from full stops, question marks, and exclamation marks, with a minimum of one.
The checks flag possible double-barrelled questions, negatives, vague time or frequency terms, and long sentences. Review each flag in context. For example, asking about price and delivery together makes it hard to answer if the experiences differ. A count over a stated period can be clearer than “often” or “recently.”
Your text stays in the browser. Reading scores and flags are review aids; they cannot establish that a survey question is valid or understood consistently. Check for leading wording and difficult recall, then pilot the question with people from your intended audience.
On plain language in public-facing questions, see Pew Research Center’s question-writing guidance.
Also while you write the questions
A useful next step
Confusing items produce straight-lining, drop-off, and answers to a different question.
Written scale labels for six common rating domains.
Turn word counts into the minutes you are asking for.
Start
Put it into practice in your next survey.
100 responses a month on Free, and no credit card. Pro is $15 a month for 10,000 responses.