Sunday, 13 January 2013

Is expert intuition good enough for conservation?


I am currently reading the excellent ‘Thinking, fast and slow by Daniel Kahneman, in this blog I will consider Daniel’s observations regarding the conditions necessary for the attainment of excellence.  Throughout the book Kahneman persuasively argues the merits of using statistical analysis to make rational decisions rather than basing decisions on intuition.  However, Kahneman also provides conditions under which expert judgement, a form of intuition, can be trusted.  Do the actions taken by conservationists meet these criteria or would individuals’ decisions improve if they were based on cold statistics instead of intuition?

The criteria:
According to Kahneman, when an expert makes decisions
·         ‘In an environment that is sufficiently regular to be predictable
·         With an opportunity to learn these regularities through prolonged experience’
‘When both of these conditions are satisfied, intuitions are likely to be skilled.’

One factor which affects the efficiency with which an expert can learn from an effect is the coupling of cause and effect.  For example, if you were learning to play chess and played a move which resulted in your queen immediately being taken then it would be fairly easy to learn from that mistake.  In contrast, if a move you made set in sequence a series of moves by you and your opponent which resulted in your queen being lost 4 moves later then it would be harder to link the cause (your initial move) and effect (loss of queen). 

How do the environments in which conservation practitioners compare to these criteria? 
Every university course in Ecology or Population Biology will draw the attention of the students to stochasticity (randomness) in determining the state of any habitat or ecosystem at any one time.  In highly random environments (e.g. those subject to a highly variable climate), if not all environments, the first criteria is violated.  For this reason experts should be wary of basing decisions on their intuition, on what they ‘feel’ to be right

Secondly, for experts to learn they must make similar decisions many times (to account for stochastic effects) with the effects of these actions closely spatially and temporally coupled (the effects must occur close by both geographically and in time) to the action.  The consequences of conservation actions such as: the planting of trees, the protection of young trees from deer and/or other browsers, reintroduction projects, habitat creation and others are not felt for many years.  Not only does this make it harder for individuals to learn from their actions, most don’t try as long term funding is sacrificed in the face of financial pressures.  Furthermore, many actions have effects which occur at a large geographical distance from their cause rendering what one intuitively feels to be right, wrong.  A good example of this comes from Chinese Lanterns which fall as debris a long way from where they are released, if the lantern fell a metre away from where it was released then I’m confident the person who released the lantern would make an effort to remove that waste (at least if it was released at home.  As it is, pollutants which cause an effect out of sight are left out of the mind of the polluter.  In order for conservation practitioners to be confident about the effects of an intervention they must be confident that no significant effects will be felt outside of the geographical area they have considered (the possible effect of badger culls increasing badger movement is one such example).

I am therefore sceptical of any ‘expert’ who sites intuition as the grounds on which they made a decision regarding decision.  As Kahneman explains the solution is objective statistical analysis.  Pool all the available data regarding the proposed conservation decision, acknowledging uncertainty.  Unfortunately, all too often the pool of available data is too small.  Conservation evidence is an organisation set up to address this problem.  It is a problem which will most easily be addressed with conservation practitioners recognising the limits of their intuition and sharing their knowledge with other practitioners with other experience.  This should not be taken as an attack on the knowledge of practitioners, rather as a comment on the irregularity of ecological systems in which effect is often spatially and/or temporally distance cause.