It's time to turn the spotlight on another PR problem: the abuse of surveys. 100 per cent of the writers in my household thought PR generated surveys were bogus.
Surveys are part of the standard PR toolkit along with press releases, case studies, launch events and buddy lunches. They are used to generate 'news' when the client hasn't anything more newsworthy to say and to give the client a platform for opionating.
And yet we find them irresistible. We want to know more about ourselves and our world and they give us (apparently) hard data. For example, I saw a pretty flimsy report once about how impatient young people were and gave it no more thought until I saw a double-page spread in The Times that was nothing more than this report with some pictures and some flannel copy.
Not only are they cheap copy filler - the writing equivalent of reality TV - but we want to believe: "Every Christmas and Easter, regular as clockwork, you can read that chocolate is good for you," says the Guardian's Bad Science blog.
Why they are bogus
- Conflict of interest. Did you ever see a press release that said '8 out of 10 cats hated our products, according to a survey'. No! PR-driven surveys always reflect the commercial interests of the sponsor. I like Paul Holmes's recent post about this phenomenon.
- Lack of transparency. You very rarely see press releases that report, in detail, on the methodology behind the survey. Often, PRs will only supply 'the full report' if you go through an elaborate mating ritual.
- Slippery statistics. Read How to Lie with Statistics (Penguin Business) for more on this.
- Poor numeracy on the part of the PR, the journalists and the public.
- Lack of context. PR-generated surveys are (by definition) one-sided. They rarely supply any context for the sruvey - reasons, causes, solutions, controversies - except for comment by the PR's client. In other words, spin.
- Public cynicism. To be fair to PRs, the government is a bigger abuser of surveys and statistics than they are. However, this just serves to discredit the whole concept for most people.
How to report surveys
- Get the whole report. If PRs try to parcel out the stats to different journalists and won't give you the whole report, don't use it.
- Report the controversy. Don't just cite the source of the survey but, if there is a potential conflict of interest, report who paid for it as well.
- Ask questions about the methodology: sample size, question rotation, response rate, who paid for it etc.
- Get numerate (do as I say, not as I do!). Good places to start: The Economist's Numbers Guide: Essentials of Business Numeracy.
- The numbers aren't the story. Use them for illustration not support and go behind the numbers to find the reasons or the surprises or the tension.
- Graphs are better than text. If you can give the information visually or find a compelling metaphor using the data, that is better than a dry recitation of the statistics. I like the way the Economist (which is essentially a numbers magazine) uses little graphs to do this. See my previous post: A picture is worth a thousand words.
- Do your own research. Kudos to the Norman Nielsen Group who do a lot of hands on research to generate their reports. Genuine, open-minded research that generates real insight is always going to be valuable.
See also: press releases
Related service: Leads