July 27, 2009
While it is important for aspiring journalists to be confident in handling and challenging numbers, of equal importance is the ability to present them in a meaningful way to a non-specialist audience – their readers.
(I’m assuming here that most of our students go on to work as general news reporters, as opposed to specialist business or science reporters).
The truism that “numbers numb” is an accurate one, just as too many acronyms can be irritating and confusing. So the first rule is to limit the number of numbers in any paragraph to three or four – and by “numbers” I mean dates, ages and house numbers as well as percentages and other data. Try this and see what I mean:
“Joe Smith, 27, of 46 Temple Road, Liverpool L8, said the 20% increase since 1996 was 10 times the national average of 23,987.”
Don’t try and cram all the numerals into a single sentence. Where possible, reduce the number of numbers by summarising them. If they are essential to the story, spread them through the story or use a bulleted list:
“Mr Smith said that, since 1996:
- The budget had increased by a fifth
- This was 10 times larger than the national average
- The national average was nearly 24,000″.
A couple of things going on here: first, fractions are better than percentages. So a 20% increase is better expressed as an increase of a fifth; a 200% increase as a tripling (why is a 200% increase a tripling rather than a doubling?).
Second, only give numerals to the degree of precision they require. The national average is 23,987 – unless the precise figure is part of the story, then that’s “nearly 24,000” in anyone’s book. Do our readers need to know, or care, what the actual figure is? (A sidenote on terminology – “nearly” or “almost” implies the true figure is slightly higher than that quoted; “more than” or “just over” should be used if the true figure is slightly lower).
This is particularly relevant when reporting on percentages, for example in council tax stories: “Liverpool council tax has gone up by 2.61%”. Now here’s an instance when the percentage figure is preferable to a fraction (“tax is up by just over a fiftieth …”), but what about the degree of precision used? Isn’t that 2.61% unnecessarily precise and wouldn’t 2.6% be just as informative? (Yes, unless the story is about the fact that the tax has increased slightly and the previous figure was exactly 2.60% ). And in this case, rounding the figure off to 3% isn’t an option, since the difference in the amount of council tax I pay at 2.6% differs significantly from the amount I pay at 3%.
That’s the key to how many figures to use – what’s sigificant to the readers? They don’t care if a national average is 23,987 or 23,982 – the figure of “nearly 24,000” tells them something informative, and they generally derive no additional information from the fuller detail. If I know my council tax has gone up 2.6%, then I know I am better off than if it had gone up by 3% – but I really don’t care if the true figure is 2.598% or 2.601%. As a reader, I gain no additional useful information from the increase in precision.
Of course, there is nothing to stop reporters from publishing the full data on their website, where readers can revel in every last decimal place; it’s just that this level of detail holds no interest for the majority of readers and therefore shouldn’t be given in print.
Rounding or approximating figures should never be done to hide laziness – “I know the true figure is around 24,000 but I can’t be bothered checking”. It is only because we’re working with numbers which are as accurate as we can get that we are justified in rounding them off when writing our stories.
As ever, summarising numerical data is an editorial judgement that relies upon our knowledge of our readership. In this, as in so much reporting that concerns numeracy, it is no different from any other type of story.