Here is something I have thought about for a long time, across twenty years of working in HR, learning, and organisational development. We spend a lot of energy analysing survey data. We spend far less time questioning the surveys themselves.
Engagement surveys. Change readiness assessments. Learning evaluations. Culture diagnostics. Leadership reviews. I have been involved in all of them, more times than I can count. And what I keep coming back to is this: the way a question is worded can quietly steer the response before the respondent has even started thinking.
That is not a flaw unique to bad surveys. It happens in well-intentioned ones too.
Survey methodology research has long shown that wording, question order, framing, and the response options we offer all influence how people interpret and answer our questions (Tourangeau, Rips and Rasinski, 2000; Dillman, Smyth and Christian, 2014).
When the design shapes the data, we stop measuring reality and start measuring our own assumptions.
Why Framing Changes Everything
Behavioural scientists have a name for this. It is called the framing effect.
Tversky and Kahneman (1981) showed that people's preferences shift depending on how information is presented, even when the underlying situation is identical. The same logic applies to the questions we ask at work.
Compare these two:
How valuable was the leadership program?
How has the leadership program impacted your work?
The first assumes value existed. The second leaves room for the person to tell you what actually happened. The difference looks small. The difference in data quality can be significant.
When the Question Has Already Decided the Answer
This one below is uncomfortable to sit with, because I have done it myself. Questions like these appear in workplace surveys constantly:
What benefits have you experienced from the new initiative?
How has the training improved your performance?
What value has the change program delivered?
Every one of those questions assumes a positive outcome has occurred. Someone who experienced no benefit, or mixed results, or genuine difficulty, is left trying to answer a question that was not really written for them.
Neutral alternatives give everyone permission to tell the truth:
What impacts, if any, has this initiative had on your work?
What benefits and challenges have you experienced since the training?
Effective consultation has to create space for all perspectives, not just the ones that confirm the program worked.
When the Response Options Do the Work Instead
Sometimes the question wording is fine, but the available responses tell a story of their own.
If every option assumes usefulness, what does a respondent do who found the initiative unhelpful? They either pick the least-wrong answer, or they abandon the survey.
Missing options I see regularly include things like: not useful, unsure, depends on the situation, or none of the above.
Survey researchers pay close attention to response scales, ordering, and the availability of options like "don't know" or "not applicable" for good reason (Tourangeau, Rips and Rasinski, 2000; OECD, 2024). It is not pedantic. It is what separates data you can trust from data that flatters the design.
The Questions We Ask About Wellbeing and Leadership
Think about how common these are in organisational surveys:
Which wellbeing initiatives do you value most?
Which leadership behaviours do you appreciate?
Which communication channels work best for you?
All useful. All incomplete. If we never ask what is not working, we collect half a picture and call it insight.
A Framework I Use: TRUST
Over the past few years, I have developed a practical check I run across survey questions before they go out. I call it the TRUST Framework, and it draws on research in behavioural science, survey methodology, and what I have learned from sitting in the middle of organisational listening for two decades.
It is not an academic model. It is a practitioner sense-check. And it works for any survey, whether you are running an engagement pulse, a change readiness assessment, a learning evaluation, or a leadership review.
T is for Test for Assumptions. Look for what the question is taking for granted. Ask yourself whether someone who had the opposite experience could still answer it comfortably. If the answer is no, rewrite it.
R is for Represent All Perspectives. If your scale goes from "very useful" to "somewhat useful," where does the person who found it useless go? Balance your options to match the full range of possible experiences.
U is for Use Neutral Language. Words like excellent, outstanding, effective, and appropriate can load a question before the respondent has a chance to form their own view. Try: "How would you describe your experience of the new onboarding program?" rather than "How effective was it?"
S is for Seek Understanding, Not Validation. The purpose of a survey is not to prove that a decision worked. It is to find out what actually happened. The most useful data often comes from the things we were not expecting to hear. Good surveys make space for surprises.
T is for Think About the Entire Journey. One neutral question in a survey full of benefit-focused questions is still a biased survey. The cumulative effect of question order and framing shapes how respondents are thinking by the time they reach your most important items. A survey is an experience, not a list.
Why This Matters for HR and L&D Professionals
We are often the people closest to organisational listening. The data we collect informs budgets, program decisions, strategic direction, and how leaders understand their people.
That is a significant responsibility.
Data is only as trustworthy as the questions used to collect it. And if those questions are quietly steering responses, the decisions that follow are built on an incomplete picture of what employees actually experienced.
I am not suggesting every survey needs to be academically perfect. In organisational life, surveys need to be practical, timely, and simple to complete. But they do need a moment of honest scrutiny.
Are we asking this neutrally? Have we left room for multiple perspectives? Could the wording already be influencing the answer? Are we trying to understand, or are we trying to validate?
The Question Behind the Question
The most powerful influence on a survey response is often not what people think. It is how the wording and structure of our questions shaped the way they were already thinking before they started.
Our job is not just to collect feedback. It is to create the conditions for honest feedback to emerge. That starts with asking better questions.
What is a survey question you have seen, or written yourself, that unintentionally shaped the response it was trying to measure?
Vanessa Trower is the Founder and Principal Consultant of Nexperk, a Sydney-based people and performance partner for large frontline workforces. She hosts The Transformation Edit and SHIFT Happens on the Front Line podcasts. Want to talk about what this could look like in your organisation? Book a chat with Nexperk.