SUNDAY RANT: A local media columnist throws a few brickbats at science. I toss a few back and give thanks for good comments.

Catching up on the New Zealand Herald website science page, which occasionally features our own work, I see that commentator Garth George recently chose to make science his rag doll for a day.

Maybe it was a slow day at the office and he simply ran out of better ideas?

Then again, looking around the ’net this doesn’t seem to be new for him. (I’m not in the habit of reading his column, so I wouldn’t know of-hand.)

But whatever.

Let’s look at what George has to say, what was said in reply, and add a few of my own thoughts.

While the title may not be his, it sweeps with a very broad brush indeed: ’Research offers little to human wisdom’ as if all research were useless.

Well done to whoever wrote the title.

Let’s consider George is only poking at human behavioural science as the first paragraph intimates.

In the opening paragraphs, I’m unsure who George is trying to slate: news editors, researchers, or both :-

Day after day in this newspaper and others we are presented with stories in which the first sentence ends with the words “a new survey shows” or “new research reveals” or “scientists have discovered”.

These items seem to have an irresistible appeal to news editors seeking urgently to fill inconvenient holes in news pages.

Few of these “research” results add anything to the sum of human knowledge for often all they do is confirm “scientifically” what we sentient humans have known either intuitively or by experience ever since we put childhood behind us.

Few would disagree with the first paragraph: tired ways of presenting science aren’t helpful. Some of the editorial choices do leave science-savvy readers wondering. The third, though, reveals a lack of understanding of why research is done.

If he is to chastise others, he ought to set a standard high enough not to be criticised himself. Or, as I love to say, you’ve got to understand what you are criticising. (I get my cue from misappropriating from Dylan’s The Times They Are A-Changin’: ’You can’t criticise what you don’t understand’.)

I’m not a fan of research that is too obvious myself. But calling research out on the basis of ’what we sentient humans have known either intuitively or by experience’ is wrong.

Intuitive notions are often wrong. To resolve different possibilities needs controlled studies, not intuition. Calling on ‘I knew that anyway’ doesn’t quite work. You also need to look at the deeper aims of the research, than taken it solely at face value. More on these later.

Let’s first look at some of what he has to say.

George cites a study he reports was described as ’when it comes to crossing the road, humans act more like lemmings than rational people when they blindly follow other pedestrians’.

First an aside, a nitpick. The only match to this exact phrase in a google search are his own words. It’s pity he doesn’t give the source if was print-only. Alternatively, if you relax the search you find near matches from many sources, but if he’s paraphrasing why then the direct quote?

But never mind that.

In reply he retorts:

So what? How many of us, cocooned in our own little mental world at the lights, have taken a step thinking the lights have changed when someone in front does, only to stop dead a split second later when we realise that the silly bugger has decided to jaywalk?

This coverage does better, linking to to the actual research – allowing readers to check it for themselves (assuming you have access) – and raising the point of it from asking the researcher: the potential relationship to ’herding, flocking and shoaling behaviors.’ (I’m not advocating the work as good or not here, just that relating the intended aim accurately might help understand why it was done – the thing George is criticising.)

As for George’s thinking: the underlying issue would seem to be not that we do it, but why.

What is it that makes us inclined to follow another’s actions? A social meme that teaches us to follow a leader? A evolutionary relic of a safety-in-numbers herd instinct?

Before you can get onto the question of why, you need to establish that it really is a established cause and effect pattern of people following others (e.g. by using social cues), not a statistical coincidence from independent events happening in similar time frames. At lot of research is about laying the foundations, determining if a particular potential causal relationship is actual rather than apparent.

Then there is his dismissal of work linking texting with other social behaviour:

It doesn’t seem to have occurred to the researchers that they have the whole thing back to front – that if a teen is bonking, or being bonked by, four or more partners then he or she is sure to be texting madly to set up assignations.

Again, I’m not interested in if the research is sound or not, but his reasoning in dismissing it.

He offers one hypothesis, as if the researchers didn’t – or couldn’t – think of it. He might ask of himself, how would he distinguish his hypothesis from that perhaps youngster more sociable in one way might be more more sociable in other ways too? Perhaps by research and statistics?

More important as one reader (Reynold, 01:57PM Friday, 19 Nov 2010) explained:

… study on texting does not assign causation, it only says there is a higher liklyhood of if X applies then Y is likly to occur, not X causes Y. Additionally, the study uses the past tense, had, when talking about sexual partners. However, Garth uses this to imply the oh so degenerative youth of today practice polyamory, instead of a series of monogamous relationships.

So, according to Reynold, not only does George display lack of understanding of what the research was doing (yet boldly going forth and criticising anyway) he has slipped a personal moral view in on the results after the fact. Good catch, I think.

This work was reported from a conference proceedings as a press release. The conference abstract for this presentation is available for those that want to read it. (I have no idea if it has been published in the peer-reviewed literature prior to the conference. If not, you might also look at it as reportage getting ahead of scientific peer review.)

It doesn’t only look at sex as readers of George’s description might think, but considers other risky behaviours and health issues such as smoking, suicide, depression, binge drinking and so on. He also leaves out that the work is comparing ‘hyper-texting’, texting more than 120 messages per school day, with these behaviours, not kids who text a few more messages than their peers. I’m left wondering if George made any effort to check what the work actually says.

I have no idea of George’s personal views on religion (although his earlier moralising might indicate his position), but his reference that ’the bible was right’ reveals another fallacy (argument from authority), most easily revealed by asking ‘how do you know it’s right?’:

That only goes to show that the Bible has it right – one man for one woman for life, and “you shall not commit adultery” – and that God didn’t proclaim such principles to spoil our fun but rather to enhance it.

The answer would be that he would not know unless be showed evidence from research – the very thing he is objecting to!

I could rattle on. It’d be too easy.

Don’t get me wrong, here. I sometimes wonder at the research reported in papers. Hell, I often do. Being excessively polite about it, some of it isn’t the best. But to dismiss research because of media reportage – not the work itself – or that he would prefer intuition over analysis, well, silly.

A key element in modern science, that commenter Melinky explains well, is that more than anecdote or ’intuitive feelings’ are needed to resolve what is right (see comment on 01:15PM Thursday, 18 Nov 2010):

Your [George’s] examples of biblical facts and sentient human “intuition” is a prime example why scientists should definitely be encouraged to back up claims of phenomena with solid research. Without the benefit of this scientific research you so disparage we would still be treating mental health issues as good old fashioned demonic possession and women would still be considered lesser beings than men. It’s common sense don’t you know?

I wonder if it has occurred to him that in criticising these reports he is unwittingly holding a mirror up to himself, that he’s doing the very thing he accuses news editors of?

As Mark K writes (02:49PM Thursday, 18 Nov 2010),

I think the author misses the point. It is often not the research that is bad but the way the media picks and chooses parts of it to suit. Just like this article really.

Understanding the aims of the research matters.

The actual aims, not what they might seem at first glance. This itself is part of the updated aims for the Ig Noble Awards, that celebrate achievements ’that first make people laugh, and then make them think.’ (My underline added.)

I think those who replied explaining the science did pretty well. Nice work.

Finally, I recommend Garth George read the Science Media Centre’s booklet Desk Guide for Covering Science.

Update: silly typo squashed.

Other articles on Code for life:

Should media only report facts and leave interpretation to the universities?

Desk Guide for Covering Science, and academic conferences

To link or not to link: mainstream media and no links at all

How long does it take you to write a science blog post?

Professor Richard Quinn responds to exam cheats