Viewport width =
October 10, 2011 | by  | in Opinion |
Share on FacebookShare on Google+Pin on PinterestTweet about this on Twitter

Failure To Communicate—Correlation does not imply causality

I’ve decided to be uncharacteristically blunt with the title of my last column of the year—Hopefully now that you all have been at university for a year or longer, you are all very aware that correlation does not imply causality.

But it is a seductive and subtle point, this one. It’s not unknown for even highly intelligent people to simply accept this in some cases without much question, and nor is it a problem that is restricted to the sciences alone. It is pervasive reminder to us all to constantly be critically thinking, first of our own ideas and then of other people’s. For when we are not, we come up with wonderful little gems like this newspaper headline I found today on stuff: “Poor skills behind Aussie pay gap.”

This was a nice little article which reported on statistical research by NZIER. The researchers had measured how many people we have employed in what kinds of industries, and how productive they were on average. It then compared those figures to the same figures for Australia, which is a really good idea to get an understanding of what may drive the pay gap between ourselves and the Australians. One of the key findings of the report was that the NZ services sector* is 25 per cent less productive than the Aussie services sector. With more than 70 per cent of all New Zealand jobs being in the services sector, the report suggests the lack of productivity in this sector is a large factor in the pay gap. So far, so good!

Now the article takes a turn for the worse, drawing from a completely different report the fact that more than one million kiwi workers have only basic qualifications or worse. It also takes from this report the fact that jobs have been lost recently in manufacturing, and that hospitality (services) is gradually taking up those employees. And now, armed with those two facts the article declares: “One million poorly educated and trained workers are the cause of a growing income gap of more than $14,000 a year between Kiwi and Australian workers, a report says.” Wow, what a leap! Nowhere in the report is the pay gap blamed on one million poorly educated and trained workers. In fact, the report explicitly states that the combination of “quality of labour, capital, and management, and regulatory environment” is of the highest priority (personally I think the poor quality of management/leadership is a bigger part of this equation than the skill of workers). And yet, by linking these two facts together and unreasonably imposing causation between them, this article has turned a wonderful and insightful report into a simple attack on low skilled workers in the public eye! If I was one of the authors of this paper, I’d be up in arms trying to defend my report, but as I am but a magazine columnist, I will turn it into a lesson: Never, ever assume causation between two correlated facts. Never.

*the report just seemed to assume the reader would know what the ‘services’ sector was, but if you are anything like me you will have been wondering exactly what ‘services’ actually is. As best as I can tell, services includes: wholesale and retail, finance, and community services. It might possibly include transport, but does not include construction, manufacturing, agriculture, utilities or mining.

Share on FacebookShare on Google+Pin on PinterestTweet about this on Twitter

About the Author ()

Comments (1)

Trackback URL / Comments RSS Feed

  1. Electrum Greenstone says:

    Appreciate your effort in trying to bring science into people’s everyday lives.

    Unfortunately, the lack of basic scientific understanding and critical analysis in many people not only continues to cloud their thinking, it is also frequently exploited by forces who stand to gain from this, e.g. business interests and politicians– forces who deflect responsibility from themselves by instead “making you afraid of it and telling you who’s to blame for it”.

    P.S. ” For instance, some years ago, research showed a clear correlation between cigarette smoking and the non-occurrence of Alzheimer’s disease.

    “It was not that smoking cigarettes somehow caused immunity from Alzheimer’s, as much as the tobacco companies might have wished, it was only that many smokers did not live long enough to get Alzheimer’s disease!

    ” Thus a co-incidence of two phenomena, even when repeated, does not mean that one phenomenon is the cause of the other.”

Recent posts

  1. An (im)possible dream: Living Wage for Vic Books
  2. Salient and VUW tussle over Official Information Act requests
  3. One Ocean
  4. Orphanage voluntourism a harmful exercise
  5. Interview with Grayson Gilmour
  6. Political Round Up
  7. A Town Like Alice — Nevil Shute
  8. Presidential Address
  9. Do You Ever Feel Like a Plastic Bag?
  10. Sport
1

Editor's Pick

In Which a Boy Leaves

: - SPONSORED - I’ve always been a fairly lucky kid. I essentially lucked out at birth, being born white, male, heterosexual, to a well off family. My life was never going to be particularly hard. And so my tale begins, with another stroke of sheer luck. After my girlfriend sugge