I love surveys.  Well, sometimes.

Regular visitors to this blog will know I’ve had the occasional go at phoney surveys.  The sort of thing carried out as a PR exercise and then dressed up as research.  They sometimes flash in front of you during TV commercials.  You know them.  Voice over – 89% of cats prefer new Tiddles.  Miniscule text at bottom of screen – Survey of 74 cats by Feline Research Associates 2006.

I’m not saying it’s phoney but the Harvard Business Review’s Daily Stat the other day (29 April) unusually brought its eager readers a statistic that wasn’t quite what it seemed.

The text is worth quoting in full.

Call-center workers retaliate against abusive callers, but rarely: Fewer than 1% of calls result in customers’ being put on hold for long periods, purposely transferred to the wrong department, or subjected to other forms of subtle sabotage, a team of researchers led by Mo Wang of the University of Maryland write in The Academy of Management Journal. By contrast, the workers are subjected to customer mistreatment in 20% to 25% of calls, the researchers say.

So far, so good.  Then the last sentence.

The data is drawn from 131 call-center employees in China.

Right.  131 people from, it turns out, one call centre serving one particular industry in one country with significant cultural differences from western society, date of survey unknown.

For anyone to make sense of Call-center workers retaliate against abusive callers, but rarely that’s the sort of contextual information even those familiar with the vagaries of statistics need let alone the average person in the street.

To be fair to the study’s authors they provide that information in their own peer-reviewed research article.

Perhaps a publication associated with one of the world’s top business schools should too.  Even for bite-sized chunks of data designed to attract attention.