The Three Common Mistakes Made in Survey Stories

News & Articles

Digital Qualitative Solutions
May 19, 2020
The Trouble with Self-Claim
May 29, 2020

The Three Common Mistakes Made in Survey Stories

Producing a publication based on a poll?

Improve your impact by avoiding these common mistakes.


Converting numbers to words

How many survey stories have you seen that read like data tables in prose format? Rattling off a long list of percentages in words is a sure - fire way to bore a reader.

A chart or graphic usually does a better job of explaining survey data than a sentence. Show the graphic - and there’s no need to repeat its contents in words.

Less is more and quality beats quantity. Most surveys intended for publication cover a lot of ground, and there’s a temptation to report everything because you can. A shorter article, cherry-picking the most powerful findings makes for a better piece than an exhaustive one that reports every question asked.

If you must share lots of data, show it in a table. Use the copy not to describe data points but to explain and share insight.

Numbers without context

Don’t report an absolute value without anything to compare. “X% of people believe Lorem ipsum”.

An absolute value is interesting only if the result is hugely counter-intuitive, or if the number is devastatingly large or small. In most surveys this won’t be the case. So what if “60% of people believe Lorem ipsum”? Is 60% good or bad, high or low? What about the other 40%?

Comparisons between cohorts are more insightful than absolute values. “X% of Group A believe Lorem ipsum, compared with Y% of Group B”. Larger sample sizes are usually associated with robustness, but the real benefit of bigger samples is more scope for analysis by subgroup.

Another way to create context is to compare results over time. This of course requires historical data to be available. If that’s not possible, a directional question can be used as a proxy. “Compared with three years ago, how do you feel about Lorem ipsum?”

Whats without the Whys

Numbers make news and therefore most research for publication is quantitative in nature. However, while quantitative research is great for measuring, it isn’t always as effective at explaining.

An insightful quote from the survey sponsor is a great way to fill the explanation gap at the same time as adding implicit branding.

Better still, use verbatims from research participants to provide context. This is straightforward to do through the inclusion of open-ended questions. Consider capturing short video clips from participants during the survey (most research data-providers offer this capability) which can then be used to create short films to be used in presentations or streamed through microsites.

Producing a publication based on a poll? Improve your impact by avoiding these common mistakes.