In two minds about clinical trials

The first way is also known as intuition or 'gut feel' and has the benefit of being incredibly fast. It's so fast, that often we don't realise we're absorbing and processing information.

This type of thinking relies on first impressions and rules of thumb, and it works well in many circumstances.

The best example I can think of is getting called out to a farm because a cow "just isn't right". If pushed, the farmer thinks a little and then says, "well, she was in later than usual".

I'd examine the cow, not really expecting to find much, and to my surprise, I'd discover a displaced abomasum or a high temperature.

The farmer used his or her observation, tallied it up against what they knew about the cow, and then realised that something wasn't right. This kind of pattern recognition happens almost unconsciously. I find that many farmers are unusually good at it.

The second way of thinking is slower and involves more conscious effort. It's the kind of thinking where you're trying to weigh up a lot of different things at once, and you're working with abstract concepts rather than visual observations.

The best example of this, I think, is herd fertility.

To get a real handle on it, you're juggling multiple figures - like six-week in-calf rate, three-week submission rate, empty rate, conception rate - and it requires sustained effort over the whole year - including heifer management, body condition, transition, heat detection, AI, etc - to improve.

When doing a fertility consultation, I always ask my farmers how they think they're going - and I often find that their gut feel doesn't match up to what the figures say.

Some farmers - although not all - get a real kick out of the numbers and enjoy mucking about with spreadsheets and cost:benefit calculations.

Others almost seem to have a sixth sense about how their cows are going and just seem to 'know' when something's wrong.

Everyone can do both types of thinking, but people are generally stronger at one than the other.

Relying on intuition has obvious benefits. It's fast and it gets the job done. It works especially well for quick feedback loops - adjusting the cows' diet to what you're seeing in the vat or picking up problems with your herd's health.

It also helps you to calibrate socially, picking up the thousands of subtle physical and verbal cues being displayed by the people you're talking to. (I'd prefer to have lunch with a group of farmers rather than scientists - apologies to my colleagues at AgriBio.)

But in abstract, counterintuitive situations, the fast way of thinking can get you into trouble.

We've already talked about the example of dairy fertility, with its long gap between receiving feedback and taking action. Another example, however, is in making scientific observations.

When trying to explain an observation we're interested in, our intuitive thinking tends to latch onto one possibility without too much provocation. Humans suffer from a phenomenon known as 'confirmation bias', where we form a hypothesis based on our observations, and then tend to selectively see evidence that confirms what we want to believe in.

This has been shown experimentally in all sorts of people and in all sorts of settings.

Everybody does it - and it's controlled by the same part of the brain that helps us make fast, rule-of-thumb decisions.

The only way to counter it is to firstly, be aware of it, and then secondly, to accurately record the data that we're using to test our claim and analyse it objectively.

The aim of this series of articles has been to get you thinking more critically about scientific trials, and to give you the tools to assess the quality of the research that you're presented with.

So, for this article, I'd like to encourage people who want to do their own clinical trials to make sure that they're recording the data instead of relying on their own gut feel.

And for those of you who are presented with other people's hypotheses, I want you to ask them if they're basing their claims on recorded evidence or anecdotal observations. If they can't show you the evidence, then it doesn't mean that they're wrong - but maybe just take their idea for now with a tiny pinch of salt.