Originally posted on December 4, 2013 @ 6:23 AM

Warped Imagination and Magical Thinking in Risk and Safety

Guest Post by Dr Rob Long from www.humandymensions.com

All the data shifting in the world is used to explain cultural change, when indeed, it is not a measure of culture.

There seems nothing more captivating to risk and safety professionals than injury data:  Canberra Times Article . It seems that any bit of data can be strained and stretched and ‘attributed’ to anything.

Hutson (The 7 Laws of Magical Thinking) calls this ‘magical thinking’. Magical thinking is most observable in fundamental attribution error  Magical thinking is appealing because it gives people a feeling of predictability and control and helps explain the unexplainable, ‘happenstance’ and coincidence. The seeking of illusory patterns and cause in information is also known as Apophenia. One of the best examples of Apophenia or fundamental attribution error was when Apple launched ‘The Shuffle’, an ipod device that played music randomly. The story of The Shuffle shows how much people are tuned to recognizing patterns and so poor at producing randomness. Listeners complained to Apple that the device would play strings and patterns of songs from the one artist, that it wasn’t really random. Apple had to introduce a new feature called ‘smart shuffle’ to allow users to manually avoid repetition. Apple announced ‘we have made the shuffle less random to make it feel more random’. Taleb wrote about this phenomena in his book Fooled by Randomness. When an organisations thinking in safety is commanded by a number, no wonder they misattribute the connection of data as cause to culture.

Coincidence is the fuel for the fire of magical thinking. We go to a sports game and see patterns in penalties or the ‘hot hand’ in shooting goals. When things are uncertain and lack predictability we attribute cause even when there is none. We seem unsatisfied with the explanation that things with humans happen randomly and that there is no deliberative or controlled cause. We use counterfactual thinking to attribute blame when there is no real connection but only randomness. We attribute the turn around in our football team’s performance to the ‘pep talk’ at half time or the substitution of a player. We find cause in wearing the ‘right’ clothes or sitting in our favourite place in the stand. Then when there is a break in the ‘pattern’ of wins or loss, we attribute all sorts of reasons as to why our team lost the game. Many people attribute such superstitious or magical thinking to cause and circumstance (http://sportsmedicine.about.com/od/sportspsychology/a/superstitions.htm ). Steve Waugh was obviously a successful cricketer because he carried a red handkerchief in his pocket. Just watch the tennis and observe the behaviour of the players, magical thinking (http://www.tennis.com.au/news/2012/07/13/friday-10-to-1-player-superstitions ).

Magical thinking or attribution gymnastics doesn’t require much effort. There is a cause to be found in any pattern or connection between data and outcome. It doesn’t matter whether the attribution is illogical or unreal, it provides comfort and that helps the attribution make sense.

One fascinating exercise in attributive gymnastics with injury data seems to be the delusional correlation of injury data to risk and safety culture. Time and time again safety people parade out injury data as if it is the measure of culture. This is certainly the case in Canberra where the Getting Them Home Safely Report and the regulator constantly use injury data to attribute safety culture causality (http://www.worksafe.act.gov.au/publication/view/1991). The report presents discussion on safety culture as if injury data describes safety culture. The data is used to attribute all kinds of connections to culture that indeed, have no logical correlation. Without knowledge of ‘Regression to the Mean’ (Kahneman Thinking Fast and Slow), and with such bias to regulatory-only thinking, the regulators find cause in injury data and obtuse time frames and then overlook a host of other cultural indicators in order to present their agenda.

What often happens as a result of magical thinking is that the regulator misattribute systems failure as culture failure or, introduces systems solutions for cultural problems. As a result supposed ‘safety culture’ surveys attribute a range of systems indicators as if they are what comprises culture. The NSW Workcover Safety Culture Survey serves as an example of this misattribution. There are so many safety culture indicators omitted from the NSW Workcover Survey, for example no assessment of:

1. Double speak

2. Induction contradictions

3. Language and discourse patterns

4. Implicit knowledge

5. Symbols and power

6. Artefacts

7. Organisational history

8. Authoritarianism and attribution

9. Denial, overconfidence, doubt

10. Perception and motivation

11. Hidden dimensions to communications

12. Subversion in subcultures,

Such misattribution of systems for culture comes from the confusion of understanding injury data and regulatory failure as culture indicators. The assessment of the dozen indicators above would give a far better understanding of culture than the ‘naïve realism’ that misattributes compliance for safety ownership. In the case of the Getting Them Home Safely Report, systems language was used synonymously with culture throughout the report as if they were one and the same. The outcome of the report has resulted in more intense systems, increased inspections, audits, penalties, on the spot fines and replicated safety data attribution yet an insidious change in the field to cultural measures such as negativity, scepticism, cynicism and double speak. So, the by-product of more systems has led to deterioration in culture rather than an improvement in culture, similar can be observed in the effects of the Office of the Federal Safety Commissioner on the culture of building and construction. All the data shifting in the world is used to explain cultural change, when indeed, it is not a measure of culture.