Blog Post

Lies, damned lies and statistics
Posted By Nick Barnes, 24/10/2014 10:30:50 AM

When conducting employee research, too often the numbers become the end game.  People become obsessed with how many people agreed with this or what proportion said that, rather than what the numbers mean and how they can be used to as a stepping stone to better understanding.

This is partly due to the emphasis researchers and statisticians put on simplifying statistics into easy-to-grasp numbers.  In doing so, we create metrics which succeed only in hiding the truth or the broader meaning behind the collective research findings.

Simple numbers are prone to manipulation, with organisations picking and choosing the statistics which will help them achieve their goal and ignoring those that don’t.  These simple, headline figures usually have consolidated totals sitting beneath them, with tallies and raw data from sample surveys further down in the process.  Most accept the headlines without digging down deeper into the numbers, where the meaning and driving patterns can often be found and explained.  And once one of these headline figures is put forward and shared, it then becomes gospel because the knowledge and / or information to challenge or refute it is unavailable.

I'm not suggesting that people apply 'creative license' to the figures; the intent of any manipulation is usually altruistic, because the people doing it are unable to understand these numbers due to the lack of exposure to the wider picture. 

And I'm not saying we shouldn't trust numbers or that statistics are largely invalid and we should rely on gut instinct.  What I am saying is that, to be reliable, objective and trustworthy, statistics need to be translated and presented, first and foremost, in a narrative. 

Numbers used holistically to construct and validate a compelling narrative of your findings will not only serve the purpose of engaging your audience, it will also avoid the construction of a finely tuned, numbers-based propaganda piece being submitted to the decision makers.

The art of developing this narrative is easier said than done, (hence the overuse of graphs and charts in most research results presentations) and requires skills which are finely tuned to interpreting data, extracting insight, and of course, communicating. 

 

For a light hearted, tongue-in-cheek look at the use / misuse of numbers to apply meaning, check out this TED Talk by Sebastian Wernicke.

 


Drop a comment
Twitter

RT @MoataTamaira: We have never needed this message more. https://t.co/YFbg2lpYLW

RT @white_owly: The one way to ensure something never becomes part of your organisational DNA is to build a department for it and fill it w…

✨ Waiting for sunrise by c1113 ✨ https://t.co/gjR6mTMKOm https://t.co/ORIdYv6q9I

Currently jumping into rabbit holes as I work on developing our next round of research projects - so many ideas and… https://t.co/OELE7nulIo

When the waiter finally arrives with your food 🤩 @DjokerNole | #AusOpen #Laureus19 https://t.co/rCmSLaoASV

Get tips on how to deliver insight-led social and online media analytical reports for your organization and/or clie… https://t.co/I7DxTgtNhQ

Get in touch!

Treble