Last week, the CIPR launched its guide for using statistics, covering why statistics are used, the meaning of statistical terms and common pitfalls, among other areas.
Now I love stats and I think they can be a valuable tool in a PR campaign. But I dislike the flippant way they can be used by the PR sector, which at times can show little respect and much recklessness in how stats are presented. As the CIPR’s CEO explains, when used well, statistics can establish the credibility and influence of a communications campaign, but when used poorly, they can damage the reputation of the organisation that used them.
In my view, the messages behind the CIPR guide are spot on; any description or interpretation of survey results must be accurate and supported by the statistics. But the guide just doesn’t go far enough in helping PRs understand how to effectively use survey results to create press releases and headlines that present the results factually and correctly.
Here are some top tips on how PRs can analyse and interpret survey results to present them effectively, but responsibly:
1. Beware the big stat headline
If a stat is so huge that it’s stating the obvious, it’s impact can be negligible. For example, if “99% of people in the UK have a mobile phone” it’s hardly a shock statistic is it. Look elsewhere in your results to find something that will take people by surprise.
2. Don’t make assumptions
Interpreting what your stats are telling you is useful, but don’t embellish your data with opinion unless it can be backed up. “Surprisingly a massive 30% of people have stopped eating chicken, suggesting that factory farming methods are putting people off.” This could make a nice message for animal rights groups, but what is there to suggest the reasoning behind the stat? And is 30% really surprising? Or massive?
3. Don’t make your stats say something significant if they don’t
I’ve lost track of the amount of times I’ve read press release headlines that try to push a story that isn’t there. Let’s think about this headline: “A massive one in four women hate chocolate”. What is this really telling us? That 75% of women simply don’t hate chocolate.
4. Big isn’t always better
Consider that in some instances, it’s the smaller stats that can make the real story. Let’s say that 5% of UK adults have admitted to stealing. If the survey has been conducted with a sample representative of the adult population, this means around 2.4m people. I’d be shocked if it was even 1% in that case.
5. Drill down into your data
Try cutting down into your survey results to uncover some new angles or find stats that strengthen your argument. You might find that a certain demographic tells a surprising story in comparison to others, giving you the hook on which to set your messaging. But be careful about drilling down too far, as the smaller sample sizes still need to be statistically relevant.
6. Strengthen with other research
Quantitative survey results can present some useful stats, but in isolation there is often little to refer to when it comes to motivation and reasoning behind the responses. Consider running qualitative research alongside your survey, maybe through focus groups or in-depth interviews. Or carry out some desktop research to further support your claims from other sources and produce a more compelling argument
7. Extrapolate for the bigger picture
If you’re surveying a truly representative sample of the population, then consider extrapolating your data to add more meaning to your results. For example, if 20% of UK households have a driveway, you can estimate that we’re talking about approximately 5m households across the UK. The latter can make this much more meaningful to the reader
Have you used survey results in PR campaigns? What tips and tricks would you recommend to others?