As the saying goes, there are three kinds of lies: lies, damned lies and statistics. The quote was attributed to Mark Twain and apparently spoken by Benjamin Disraeli (a 19th century Prime Minister). Though as you can see here, the phrase has quite a history. But, I digress. This week’s Wicked Wednesday prompt is about statistics. A topic I’m really interested in, because of what data can tell us, how they are manipulated and used to inform and fool.
Statistics as a branch of maths found me when I was 16. Maths was far from my favourite subject, and I wasn’t very good at it. So, I wasn’t that surprised to fail my O level and need to retake it in the 6th form. I was offered the chance to do my qualification in statistics instead of retaking maths. I thought this was worth doing as it might be easier to concentrate on one area. It was, but also it was fascinating and proved useful once I went into nursing. Healthcare produces a lot of data and statistics about that data is churned out at the rate of knots. Actually understanding what the mean is and how standard deviation is measured has been really useful. Both my degrees required me to examine quantitative and qualitative research so, I’ve never stopped using statistics since.
We are all a statistic or 500
The trouble with statistics is that it can tell you pretty much anything you want it to say. Data about us is collected from before we are born. As soon as someone is pregnant they enter a system of data collection. Birth, immunisations, development checks, attendance at play group and so it goes on. Back in the day, little was recorded about us save the statutory things. But the advent of computers, digitalisation and mobile phones means thousands of items of data are collected on us every year. Decisions are made by politicians, companies, providers of healthcare and schools etc. based on data collected about us.
For example, NHS England publish have a page that signposts you to all the statistical reports they produce. The Office for National Statistics publish all of the National data for England and Wales on births, deaths, employment etc. Ofcom publish information about mobile phone usage and how it has changed. No doubt the information is gathered through information supplied by phone operators. This page also refers to some research carried out. Though it says nothing about the methodology, or even how many people took part. That of course is the problem.
Reporting of statistics
Headlines are the thing. We live an era where attention span is short. This article refers to a research study on this topic. The article itself is a few hundred words long and of course the author has pulled out the salient points; there is just too much information out there. We read something then move on quickly to the next thing. True for some, but not for others I’m sure. Trouble is, you can never be sure a journalist has read the whole research report or understood it.
When the ONS tell us how many deaths have occurred this week, we can trust that the person doing the reporting knows their subject. But as soon as that report is transcribed into an article or mentioned on TV someone has decided they need to explain it to you. Covid-19 has led to the publication of (probably) hundreds of thousands pieces of statistical information world wide. Charts that show infection rates and deaths are produced by multiple organisations. Depending on how they are displayed they look different.
Whole numbers are great but its difficult to compare a huge country with a small one. Or huge urban areas with a rural one of the same size. People with certain health conditions are more at risk as are men and people from BME communities. Or so we are told. For example it has been reported that people with type 1 diabetes are at greater risk than those with type 2. But age and sex were also factors, so it isn’t as straight forward as the article suggests. Trouble is, to understand what is really being said you have to go to the actual research and most people don’t have the time inclination or indeed attention span to do so.
Which leads us to the problem of having too much data and being interested by it. Whole hours and indeed days can be lost to statistics if you aren’t careful. Myth busting articles written by people payed to do so are useful, but that doesn’t mean I don’t go looking for more information. Master is the same, so much so he has been collating his own data on Covid throughout the pandemic. I guess we are both data and statistics nerds. Also we tend not to believe a lot of what is written about them. I’ll be glad when there are new TV shows, music or holiday destinations to read about instead. Then I’ll maybe give my tired brain a chance to recover!