Our Story
Surveys were once a bastion of scientific rigour
Forty years ago, survey research was a discipline dominated by trained specialists who understood the nuanced interplay of sampling, probability, statistics, behavioural psychology, and questionnaire design.
Today, the landscape is drastically different. The rise of online platforms has democratized survey creation… to its detriment. Platforms like SurveyMonkey have hosted over 150 million surveys in the last two decades alone, replacing scientific methodology with simplistic DIY tools. Most DIY survey creators and the new insta-expert survey suppliers are unaware that there even is a science behind survey research.
The decline doesn’t stop at survey creation. The analysis of survey data has suffered similarly. Training in the process of data analysis is often absent, while modern software allows users to press a button to spit out statistics with no grounding in the theories that underpin their appropriate use and interpretation.
Restoring scientific foundations in a fake science era
During the COVID pandemic, the stakes of “fake science” became painfully clear. As the founder of the Survey Science Institute, I observe with alarm a world retreating from the principles of the Scientific Revolution across a wide range of contexts and fields. With surveys, many DIYers and untrained suppliers are even unaware that accurate data is not an automatic outcome of asking questions and getting responses.
From flaws to fixes
The vast majority of today’s surveys generate flawed data. Inaccurate survey data and analytics directly impact decisions made by businesses, governments, and organizations worldwide. Whether it’s your in-house DIY project, a committee-driven process, or your low-cost supplier, the lack of awareness of survey methodology is producing misleading data that steer your critical decisions.
Our Story
A shift, not a decline
Survey research has changed. It was once a discipline dominated by trained specialists — professionals who understood the interplay of sampling, probability, statistics, behavioural psychology, and questionnaire design. Today, accessible platforms such as SurveyMonkey and Google Forms, combined with click-button statistical software, have made survey creation available to anyone.
This is not a story about carelessness, or about people rejecting science. Most DIY survey creators — and a surprising number of professional suppliers — are not skeptical of survey science. They simply have no idea it exists. Nobody told them there was a discipline to learn in the first place.
No blueprint, no foundation
The same is true of analysis. Modern software allows users to generate statistics at the press of a button, with no grounding in the theories that determine whether those results are valid, or when they are not. The tools got easier. The understanding required to use them well did not.
Bad data looks the same as good data
This is what makes the gap consequential rather than simply unfortunate: it is invisible. A flawed survey carries no warning label. It looks exactly as credible as a rigorous one, until the decisions built upon it begin to fail.
Where I come in
During the COVID pandemic, I watched something else unfold — a different kind of erosion, where people actively rejected the authority of science itself. It was a stark contrast to what I had spent my career observing in survey research, where the issue was never rejection. It was absence. People were not turning away from a discipline they knew and distrusted. They were operating in a space where that discipline had simply never been introduced to them.
That distinction is what led me to found the Survey Science Institute — and to begin writing a book on the subject. Accurate data is not an automatic outcome of asking questions and collecting responses. It is the product of a discipline most people have never had the chance to learn. Most organizations making decisions based on survey results have no reliable way of knowing whether theirs will hold up.