Bertha Harian

So you see a number in a headline. Do you ask yourself why the number is there? Is it because it is a very big or small or surprising number? But what is “big’’ to you may not be “big’’ to someone in the know. Or it might look “small’’ to you but it’s significant for being publicised at all!

Stats are facts, but not necessarily the whole truth.

Numbers only make sense in a context, unless it’s the first time ever we see a number, like the number of Chinese nationals who are now citizens. Usually the news is about a change in the numbers. Over time. Over space. Or it’s about the difference in the rate of change. Faster or slower than before.

Statistics underpin most news reports because they seem to be the most reliable evidence of change. But change can be insignificant as well. It’s not news if more old people have hospital operations given that we are an ageing population. It’s news if fewer of them do so or if the numbers are more than what would be a normal growth rate. So think harder when you come across an M&M (more and more) story and don’t get all worked up too soon.

The most popular statistics in local media are the ones that are actually the most difficult to fathom. They are mainly economic, like GDP numbers, non-oil domestic exports and trade/investment flows. They look like numbers which only businessmen and economists should care about although they underpin our economic health and wealth. Most people are probably more concerned with the Consumer Price Index because it is an indication of cost of living. Put it as year-on-year change, however, and people start disbelieving numbers because they can’t remember what things were like last year. But they can remember last month.

The choice of numbers is dependent on the significance of the changes, which could be in the form of preliminary estimates and predictions as well. There is also the frequency with which the numbers are pushed out, whether every month, every quarter or yearly. That’s why sometimes reports on numbers seem so familiar and yawn-inducing. You feel like you’ve read them before. Only the numbers have changed.

A good journalist will put the numbers in perspective or pick the numbers which would interest the most number (sorry!) of people. It’s not unusual to read different news reports emanating from the same set of numbers. Much depends on the audience the media caters for, and the editorial imperative. So a business news media might focus on, say, the number of luxury cars sold while a general interest media would go with sales of Toyotas.

Of course numbers can deceive. You can choose to highlight only positive numbers and ignore the negative ones. You can talk accurately about a 10-fold rise in revenue – and hopes that nobody asks you about the embarrassingly small base. Or pick a timeline that makes you look good – and skip all the times you look bad. Or boast about how your business is faring now compared to 2020, instead of using the more stable pre-Covid period.

Did you know that GST actually went up by 28 per cent over the two years? This is accurate as a rise from 7 to 9 per cent. But we prefer to say that it went up by 2 percentage points. That’s less scary.

I have always wondered why the media calls an increase a hike sometimes, and a slight uptick in others. But I know the sentiments that both phrases can arouse on reading. Likewise some stats have been described as a plunge instead of a dip, each laden with meaning. Is this spin or analysis? That depends on the, err, depth of knowledge of the writer over the subject.

The word I dislike is ‘moderate’ or ‘moderation’. This could refer to numbers going up OR down but the net effect is a perception of normality or stability.

What about the results of surveys? They seem to be a big deal in Singapore media which seem to lap them up with little question going by its preponderance. Most times, we must assume that the media have done the due diligence in ensuring that the survey is correctly done and without bias on the part of the sponsor.

The usual thing to do is to let readers know a little about how the survey was done, including sample size, statistical significance and how questions are asked. This is sometimes ignored in the rush to report the results of the survey. Which is why surveys are so easily attacked, especially when its mechanics are not properly explained.

Just give us the charts you say? I also say but even charts and graphs need words. It should be made clear what the chart/graphs are trying to show the reader, that is, it must have a headline that gives the news rather than a blunt X versus Y axis label.