Wednesday, May 22, 2024

Why You Need To Learn To Love Statistics.

March 11, 2021 by  
Filed under Money/Business, Tech/Internet, Weekly Columns

( Pick up any news article and there’s a solid chance you’ll find statistics in it somewhere. This is particularly likely if it’s on politics or economics. Statistics can be massively useful, but they can also be plain misinformation. If you want to be able to tell truth from deception, you need a basic understanding of statistics. Here’s a quick guide to the basics.

You need a source

Any time someone quotes a statistic, they should also quote the source. If they don’t then it’s an immediate red flag. The person may be speaking in complete good faith. They may, however, simply have got their facts wrong, or been misled themselves.

If you’re given the source of a statistic, then check it yourself. Firstly, you want to know who carried out the study? Was it a reputable professional like biostatistician Dr. Dorian Wilkerson or was it someone more questionable like a lobbying organization?  

Secondly, you want to know who funded the study. They shouldn’t have influenced its conclusion. You should, however, be aware of the possibility that they did.

You need to understand averages

Whenever you read the term “the average”, your immediate response should be “which average”? There are three ways to calculate an average and they can give very different results.

The mean is the average you get when you take all the data values and divide the total by the number of data values.  

The median is the average you get when you list your data values in order and look for the number in the middle. To do this, you divide the total number of data values by two. If the result is an odd number, the value in this position is the median. If it’s even, you add together the two middle numbers and divide by two.

The mode is the average you get if you use the most common data value to represent the whole.  

Averages in the real world

Say you want to figure out the average wage for your area. You poll 10 people and discover their hourly wages are $10, $10, $10, $10, $15, $15, $15, $20, $100 and $500. So what’s the average? At the mean, it’s $70.5. At the median, it’s $15. At the mode, it’s $10.

All three of those results are mathematically-valid. They do, however, all give very different impressions. For example, if you were campaigning for a wage rise, you’d quote the average at the median or mode. If, however, you were campaigning against it, you’d quote the average at the mean.

You need to know the context

There are three key facts you need to know about a statistic before you can decide whether you can take it seriously. Firstly, you need to know how big it was. In statistics, big really is beautiful, or at least potentially more accurate.

Secondly, you need to know how the sampling was done. Who or what was surveyed, where, and in what way?

Thirdly, you need to know when the study was undertaken. The older it is, the less likely it is to be valid today, even if it was once a great piece of research.

Staff Writer; Joe Adams

Speak Your Mind

Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!