Statistics — most people go into a spin when they hear this word. 

Statistics are all around us, however they are used all too often to convince (or manipulate) the public.  Those in government use the numbers to tells us things are improving (this is particularly the case with economic statistics), whilst the opposition uses the same numbers to tell us things are going bad, downhill. Others use the data to convince us that we need to change — the way we think is no longer the norm and it is time to make some changes for the common good of the country.  

So who is telling the truth? 

The problem is, we live in a society that is not statistically literate.  Therefore many people would not have a clue if the statistics they are being fed are accurately or not.   This is particularly the case during elections — beware of any statistics you hear, they are used to manipulate the truth and both sides do it — in all countries.

All of this gives statistics are very bad name, for me, they tell  stories about ourselves —illustrating the changes in society over time, they tell us if circumstances are improving, deteriorating or staying much same, they gauge the mood of the nation.  Statistics are fascinating and we should take far more noticed of them.  Perhaps then, we will know if they are being used against us.

The other issue with statistics is the collection of data in the first place. Questionnaires can be skewed giving  the outcome required — who would know if this had happened, does anyone care? Small samples are often used (running a survey is expensive) and smaller the sample the less likely it will represent the views of the population creating useless data, but still published in newspapers anyway — the public don’t have a clue.

Here is a perfect example of using statistics to influence government policy — recently the newspapers have been filled with statistics illustrating the changing support for gay marriages — “the majority (62%) of Australians agree that same-sex couples should be able to marry up from 60% in 2009” (Galaxy poll).   This is what is splashed across the papers — however if you look closer at the report only 25% “strongly agree” with another 21% who “agree” whilst another 21% “strongly disagree”, 33% sit somewhere in the middle.    These are, of course, aimed at the government in an attempt to lobby for changes to the Marriage Act.   And, the sample size was just over 1000 people with their responds weighted to represent the entire population of Australia.  Maybe if a larger sample was done the results would be similar, as it hasn’t, it is hard to tell.

It’s not that the statistics are evil, it is how they are used that is the problem.

My message is, teach your children statistics — how to read and understand them.  This way they will have a better understanding of the how they are being used and being manipulated.  If you see statistics in the newspaper, go back to the source, it isn’t hard to find out more about the data, check out the methodology.   Please don’t get me wrong, not all data is poorly collected, those done by the Australian Bureau of Statistics (our official statistical agency) are very sound, however it is when they are used, it becomes a problem. 



  1. Jo,

    You make some interesting points in regards statistical information. It actually doesn't surprise me that statistical information can be misused to satisfy someone's needs, rather they be good or bad.

    However, you are correct that we should try and learn something about this subject so that we are not easily "led astray" simply by what people try to convey to us.


    -Lady Rose

    P.S. I remember taking a Statistics course in college. It was not my favorite subject. :)

  2. I find statistics very interesting too. The thing that I find fascinating is the fact that I am very rarely asked for my opinion on the subjects the statistics are based on. How many people you know are asked questions? Who actually answers these surveys - are they a true cross-cut of society, or are they specific groups of people who like to answer such surveys and thus the surveys only represent a particular kind of person??? I don't know a lot about how they choose people to do the surveys, but this is something I have wondered about. I really enjoy surveys and so I have tried joining organisations who send out surveys, and what they do is ask you a range of questions and then decide (based on your answers) whether you should do their surveys... and so that is why I ask the question. It is not a good representation of society if they cut certain people out because they answer a question wrong on a "get-to-know-you" questionaire!!

  3. Yes I read that recent news report too and was wondering where they did that survey!
    Good points Jo to educate our kids on this subject.

  4. Clara - the official statistics by the ABS are done very reliable and the methodology is sound. They sample large numbers of people (eg for the Labour Force Survey) and obtain a representative sample through mathematical formulas. They also get around 94% response rate (which is very high - meaning almost everyone survey answers).

    However when a survey only samples 1000 people within a population of 22 million it is very difficult to obtain the same representative sample, however almost all polling is done with that size. And if 20% do not respond, the sample drops to 800 people.

    The problem with large samples is cost - it costs many $$ to run a large survey, some are well into the millions. Small survey's are cheaper.

    Sometimes a person is removed becuase they already have enough of that demographic eg young male. If the survey is health related, they may only want people with a particular condition. Therefore they will only be representing a portion of the population.

  5. And of course the answers are markedly different between country & city folks, country people generally being far more conservative in their views & far less represented in these sort of things. They are always skewered. I pay no attention to them. You can make stats say whatever you want. The one you quoted from this week's paper is a classic. It's as good as a lie ~ yet people will believe it because they read it in the paper & it's *statistics*!


Post a Comment