There is a truism about working with numbers namely that "numbers do not lie." This sounds as if they have got a magical property to extract truth out of anything: separating fact from fiction.
Perhaps the reason why we have confidence in numbers like this is that, mathematics by itself is based on some laws which, if violated, produce undesired outcomes.
But there is something else with numbers that give them more force than this. This is why perhaps Huff1 cautions against over-reliance on numbers without understanding their context, or even better knowing statistics.
We are used to "1 + 1" being 2, and not any other number in the universe: so we know that owing to the laws of mathematics, that are based on the physical world, if we add "1" element to another "1" element, we should have "2" elements as a result.
But do numbers lie or tell the truth?
I don't think they do either of the two: numbers just state a result of an action. that is all.
What this simply means is that, if you erroneously supply numbers and expect a correct outcome, this would fail.
Think of it this way: gigo, that is, "Garbage in, Garbage Out" is the rule of numbers;
Use bad input, and you get bad output;
Conversely, put good input to get good output.
This is a mathematical function: whatever you input into the
function, it is processed as it is and you get its output.
Perhaps, this is why they say numbers do not lie after all: they just tell you what they have, whether good or bad.
But it means they can be manipulated: you can't rely on "mere numbers" to tell you whether the story behind those numbers indeed happened. Which brings me to the point of the post:
The power of numbers
They thought there was no one with it, because no report had been made to that effect.
But as soon as the figures started trickling in, there was not only realisation, but a panic. Indeed we were in danger of contracting it, and worse still, we were not sure who the next casualty would be.
This is what figures do to a people: they call them up to action.
This is because they put every event in measurable units. They rep represent each event and put it on a scale.
Numbers, due to their property of quantifying things, make it possible to measure anything.
But as this article shows, the figures we are getting may not really be real figures: they are just official figures. Which means that the real ones may even be higher.
Official figures often reflect how much
is being done at a national level.
So we may have this layout for any event:
|Event||Real Stat||Official STat||Difference|
|Road Accidents per annum||153,000||92,000||-61,000|
Well, this is just fictitious data: but as you can see, when you compare differences between the second and the third columns, you see that in some cases, there are big differences between the official and actual figures.
The above table shows these differences in the Fourth column in this manner:
When the official position inflates, a Positive "+" symbol is added; while
there is a reduction, a negative sign (-) is used.
The pattern in the above table is interesting in that:
Data that tends to put a country in better light, for instance its education or employment rate, tends to be magnified. (See Rows 1 and 2).
That which tends to put the country in bad light, for instance number of road accidents or school dropouts, is reduced. (see Rows 3 and 4).
But does this mean we have to distrust every official data we get? No. Actually, we have to rely on it. It gives us somewhere to start.
We can make some interesting analysis based on that data. In any case, as long as we do not have real data, we have to stick with what we have for the moment.
And, this trend of exaggerating attractive data and reducing damaging information is not unique to any one country. It seems as if a number of governments do that.
But why do we trust these statistics anyway?
Because we have no way to dispute them. The only way we can dispute numbers presented to us is to produce another set of numbers and prove that ours, and not the reported, figures represent the truth. Huh.
This is the quandary of electoral petitions: if one has to dispute the reported figures, they have to do so not only by presenting another set of figures, but even more importantly, by showing irregularities in the conduct of the elections themselves.
In the case Chamisa v Mnangagwa & 24 Others (CCZ 42/18) 2018, the lawyer for the applicant was correctly putting across the statement of "numbers do not lie". However, his case fell simply on the applicant failing to produce any solid statistics to contradict those numbers presented officially. This could simply have been done by just furnishing the court with an original Form where primary data was entered at the polling stations.3
Another reason why we have to stick with official position is being pragmatic: We all know it from research.
There are many places where data may be wrongly entered. Either
during its collection, entry or even reporting.
Points are proven or Challenged by Numbers
Well, there it is. Numbers summarise a long story into statistics. This is important in preparing reports or proposals:
People may not have the whole day to hear the greatness of your
research, instead they just need it expressed in statistics.
These give people a rough idea of how much resources are needed to
have the research carried out, how different the hypothesised statistic from the true population parameter, and the time it will take to complete this research.
perhaps, this, more than anything else, gives quantitative research
an appeal in a number of disciplines.
The reported figures of Covid-19 today, are they a true reflection of how exposed we are, how people are now increasingly contracting COVID-19 or something else?
The answer could be anything, but numbers help present a disaster in its proper magnitude. Until we are persuaded by how big or small something is, we may not have an idea on how to act.
However, this fact of numbers used as a tool to convey a message and influence decision-making is a double-edged sword. It can be manipulated to meet other nefarious ends, as Huff puts it in his famous book, How to Lie with Statistics.1