There’s No Such Thing As Average: an article by: GF Willmetts.
Do you find it odd how much we rely on statistical terms like ‘average,’ ‘norm,’ ‘normal,’ or ‘mean’? Different words, but they represent the same concept—the middle of a statistical line, implying that anything else is less than perfect when, in reality, no one perfectly meets that standard. It’s probably a human tendency to pour all comparable data into a statistical pile, ignore the extremes on each side, and select something in the middle as if it’s representative. This might be acceptable in number statistics, but in reality, it doesn’t necessarily reflect, well… reality. You might determine an average height or weight, but it’s unlikely you’ll find an ‘average person’ who matches all the statistics available. Who wants to be average? We all like to be better if we can. Even IQ has been divided into various aspects now, rather than just assessing someone who is good at answering IQ questions, which can be improved upon with practice. Statistics can teach us things that don’t quite match up with reality.
The desire to reduce things to numbers can be interpreted in many ways. It can make people feel inferior or superior to others, affecting how they are treated in life. When it comes to intelligence, you don’t really need numbers to show if someone is smart or dim, as a lot of the time, it’s self-evident. It makes some people feel superior or inferior, but it becomes more complex depending on how benevolent or malevolent we or they are. In the past, geeks were often looked down on, but in the computer age, we have become a valuable commodity. One-upmanship tends to be another human fallacy, although it often means someone, somewhere, will feel like they are at the bottom of the pile. The recognition that there are different types of intelligence has meant we can all be smart in different ways.
Different Types of Statistics
One thing you shouldn’t do is think all statistics are the same. Physical data, like measurements, are clearly not the same as opinions or personal choices, although there is some connection if you’re comparing them. You only have to look at the final questions asking your age and employment status—they are looking for correlations between your choices. Some companies even ask those questions first, and if you don’t fit the criteria they’re after, you get dropped on the spot. So you have to ask what kind of representation they want, and it makes no allowance for the fact that some people might actually lie. After all, even a negative result will provide some representation of balance. It might save people’s time, but it also means an imbalance in bias. If you ever encounter a doorstep or phone statistician who isn’t prepared for answers that might not align with what they want, then avoid their results. I had one last year about my listening habits on the radio. I don’t listen to the radio, but she was happy to record my answers because I was at one extreme.
Physical statistics come into their own when you want, say, a height order, as with mountain heights or setting a world record. You probably get a nice drop of Guinness from that, even if it’s for odd extremes. Of course, this presumes all measurements are accurate and taken the same way. Even so, a snowy peak might possibly hide the true height of a mountain, but it only becomes important if the heights of similarly sized mountains are within inches of each other. That’s an odd sort of statistic, and people rarely question such records.
A lot of the time, statistics tend to come up more with population data—how it’s divided by sex, caste, and age, and how the census compares to previous ones to track population growth. All of this is rather handy for working out the needs for schools, jobs, and such, ensuring there might be enough for all. Statistics do actually work for the community but get shaded when people think all statistics are lies. It’s the interpretations you have to worry about.
Living Up to Statistics
When it comes to physical statistics, there are some things that are harder to change. You can’t really change your height because it’s based on your bone lengths. Weight is different, although the new drugs that can change it suggest it might not be under our control but rather an odd hormone mix. We know a heavy weight isn’t good for the heart, but is there an average that can be maintained for long without a steady diet? In the long term, if the drug becomes commonly used, then weight statistics are bound to change quite significantly. That’s a moot point. Even so, the so-called averages can vary from country to country, obviously indicating that an ‘average’ is in the eye of the beholder, not something to try to attain. There is no overall ‘normal’; a lot of people just like to think there is. For us geeks, I doubt we’d ever qualify as ‘normal’. It wouldn’t feel right.
What it really means is that although some statistics are just data, some can be useful—but not all. The fact that we are encouraged to heed statistics doesn’t mean they are all accurate or truly representative in their interpretation. Geek questions will identify our range of knowledge in various subjects, but finding a norm in geekdom is a lot harder because not all of us have the same talents or interests. For us here, we’re SF-orientated geeks, but would a sports geek be comparable to us?
Statistical Trends
Statistics of any sort indicate a trend that can vary with time. Our ancestors were much smaller than we are, so should we really compare that with modern-day heights? Different times and changes in diet have caused some refinements—or even worse—depending on how you look at it. They are only an indication, not something that needs to be followed religiously.
Political Statistics
Just because a graph shows a likely win, even when it comes to voting, doesn’t make it so. Statisticians can get seriously unstuck if their ‘average’ population isn’t a wide enough representation. It’s a prediction, not a fact until it happens. Unless a vote is taken and everyone participates, the prediction is meaningless. The wrong cross-section in the 2016 UK General Election led to an inaccurate prediction. Getting a true representation in a small number is impossible when the UK has 68 million people, let alone in even bigger countries. In other words, statistics is not an exact science and not really a science at all when it comes to opinions. It doesn’t allow for people to lie or even to abstain if they haven’t made up their minds yet. The ‘Don’t Know’ column then becomes a very grey, undecided number. If anything, it needs more qualification as to what they don’t know.
Do You Know What Statistics Is?
Considering how many people are, shall we say, not mathematically inclined and don’t even understand statistics but still use its numbers, it does sound paradoxical. It’s easily tested by asking if they know what a standard deviation is. That’s a small percentage either side of the average that makes it a less precise number but allows for error. So why do so many people rely on it unless they see it as an easy number? Mostly because the human herd instinct kicks in regarding what might be seen as allowable and acceptable when really, it’s only a number. Then again, how many people think numbers have some sort of magical properties, as witnessed by those who pick the same numbers all the time in gambling games.
Statistics Can Tell Untruths
That’s not to say statistics can’t be manipulated. The size and variety of the sample taken work better with larger representation. If you select only a particular part of the population, based on jobs or status, then you automatically get a bias. A question can be written in such a way as to direct a certain response, meaning that a ‘no’ can have a different bias under certain circumstances. If you say ‘Don’t know,’ then you’re also misdirecting the answer because it leaves more weight to those who’ve given a definitive response. It’s a tricky call, isn’t it?
The choice of questions can also manipulate and gear a response toward what the sponsor wants. I did say sponsor. A lot of statistics are taken for particular companies and sponsors rather than being unbiased and can be swayed toward what they want. This has a lot more significance on the ‘Don’t know’ category because it often means the respondents probably didn’t have an opinion on the question, just no answer that fit it. When looking at statistics, always consider the above before making a decision based on them. A particularly worded question might not give you the chance to provide the answer you really want to give.
How to Read Statistics
That’s not to say statistics can’t be useful, but it does make sense to understand how to read them. You don’t need a degree. As noted above, it’s more a question of looking at the representation and the questions asked. If the information isn’t readily available, then look elsewhere. Ask yourself how you would answer the same questions and see where that would place you on the statistical curve. Chances are, you wouldn’t necessarily be in the average, but leaning one way or the other will at least reassure you that you are not alone in your opinion.
Statistics represent so much—from opinion to physical data—that they can’t all be read in the same way.
It’s been said that you should take statistics with a pinch of salt, but really, you need to understand who is asking for the information and what they really want. Don’t take the information as a given; just consider it a sample of what is out there. If 9 out of 10 agree on something, what did the other one think, if not a ‘don’t know’? Statistics reveal trends far more than facts.
© GF Willmetts 2024
All rights reserved
Ask before borrowing
