Hi Nicola,

It's not an HR-specific podcast, but the one I've found most useful at work is Radio 4's 'More or Less: Behind the Statistics'. It's really honed my ability to question and use statistical information of any kind - particularly useful in today's data-driven HR world. They take headline statistics of various kinds and thoroughly interrogate the data behind it. Recent ones have included: what's the increase in performance depending on where a cyclist rides in the peloton, counting rough sleepers, and are there really more stars in the universe than all the grains of sand on the beaches of the world.

One of their articles, an old one but one that still tickles me, was looking at a Telegraph heading from 2015 that said 1 in 5 passengers admit to cheating on their partners whilst on a cruise, and that in 80% of these cases their partner was on board. The article continued with salacious assumptions on what was happening in these floating sex dens, in the way only papers can.

To cut a long story short...

Behind the Statistics showed that the survey from which this was taken had replies from 300 people, which amounts to 0.0002% of the annual world-wide cruise numbers. If all of these had agreed, it's still a statistically insignificant number.

This example is useful to remember when being given % data - percentage of what?

They also showed that the questions posed were unintentionally biased. For example, the 1 in 5 number the Telegraph referred to was from question 2: 'Was it [sex] with someone other than your husband/wife/partner', to which 17% had said 'yes'. But you could have answered yes if you were single, or had remarried but been on cruises with your previous spouse, or currently be separated, etc. Other questions were also phrased badly.

This example is useful to remember when compiling questionnaires, remember to look at the questions from all directions in order to minimise bias.

I use this Behind the Statistics example a lot when training to show how statistics can be misread or misrepresented. Get people to talk about the headline first, then introduce the questions and discuss the answers, then look at the problems with what's been said. I've attached a copy of my notes if anyone else is interested.