Don’t trust your computer

A researcher at Leeds University Business School has found that people who use computers to help them make good decisions are often unwittingly being led into making bad ones.

John Maule, a professor of human decision making and the director of the Centre for Decision Research at Leeds University, said that if someone has a hunch about how they should take a particular decision or solve a problem, they tend to look for anything that confirms what they initially believe to be true. This can make them even more confident about what they are doing, even when it is wrong.

A second problem, Prof Maule says, occurs when people use computers to analyse and evaluate large quantities of information. He believes that there are situations when it is better to follow your intuitions instead of using a computer. The trick, he says, is to know when to turn off the computer and go with your gut instinct – a trait that Maule claims that most decision makers are not taught.

Individuals and organisations spend huge quantities of money on computer systems that are supposed to help them make better decisions – and in some cases this is justified. However, to prevent the problems caused by mental ’short cuts’, a major re-think in how we use computers is needed, Prof Maule believes.

He said people should be trained to think more critically about what computers are advising them to do and called for the development of a new wave of more sophisticated IT systems that are based on a better understanding of human psychology.

I can’t agree more. I could cite several cases I have come across where engineers have made the mistake of believing that because they have developed one solution to an engineering problem – and then confirmed its validity on a computer – that the solution is then clearly the most advantageous one, when, in fact, it later proves not to be.

I have also encountered cases where engineers have used what have been marketed as ‘sophisticated computer spreadsheets’ to analyse reams of data using complex mathematical formulae, implicitly believing the results that were presented to them, only to find that when their superior checked the data that it was found to be erroneous – with the spreadsheet software at fault.

So while computers have undoubtedly proved extremely beneficial to both engineers and scientists, they also have the ability to make us lazy and too reliant upon them.

Until they do indeed become sophisticated enough to be trusted explicitly – which may be quite a while from now – it’s still good to be wary of any results that they present us with.

Dave Wilson
Editor, Electronicstalk

Dave’s comments form part of the weekly Electronicstalk newsletter, which also includes a round-up of the latest electronic products and services for engineers. To subscribe click here