It’s a digital world: use digits for numbers

 Usability rules when writing numbers in text

What if everything you ever learned about spelling out numbers was useless?

I strongly believe that we should all use digits for numbers, regardless of whether we are writing for online or for other media and of whether we are writing small numbers or big ones. The old conventions about spelling out numbers are silly, obsolete and … well, just plain dumb.

In general, English language and grammar rules are helpful. As a writer, I find that most of those rules have a clear logic behind them and that knowing and following the rules helps me write texts that are easier to read and understand.

But the generally-accepted rules about numbers are inconsistent, unclear and unhelpful.

The rules about spelling out numbers are arbitrary and inconsistent

A 5-minute search will prove to you that there is no consensus in English on how to write numbers in text. In fact, it all seems rather arbitrary, if not downright useless. Styleguides disagree on:

  • which numbers should be spelled out: 1-9, 1-10, 1-20, or even 1-99; round numbers like 20 or 50; only multiples of 100; etc.
  • whether to write three hundred thousand, 300 thousand or 300,000
  • whether or not you should spell out numbers at the start of a sentence.

Meanwhile, usability experts recommend using digits for all numbers, claiming that they stand out more easily and the brain processes them more quickly.

So which advice should writers follow?

Usability RULES!

It seems to me that today, when most of us read more on screens than on paper (regardless of the channel the text was originally created for), the answer is obvious. We have so much information to process in a day, why slow down our reading just to obey some arbitrary and archaic rules?

Moreover, if digits are easier to read than numbers spelled out, and the old rules are all arbitrary anyway, why restrict this great usability trick to texts intended for reading on-screen?

So I believe the time has come to let go of the old rules and start writing for today’s readers: use digits for numbers, always.

It’s just a matter of time: 1 day we’ll be used to it *.

Well, that’s just being provocative. You could argue that “1 day” means “some/any day” rather than a specific quantity of days and you should therefore use the word one rather than the digit 1. Usability experts also state that non-specific numbers can be spelled out.

Sometimes, the best rule of all is use your common sense.

top of page

7 comments

So happy to see this! I agree absolutely. Digits over numbers wins the debate. Let’s get into the 21st century.

Hi Ramona, thanks for the support! I think we’re in a minority with this one but we’ll get there someday 🙂

Dick Margulis

This begs for a well-designed experiment.

Many years ago (more than fifty, actually), the Starch Organization did A/B testing on this proposition in the context of a daily newspaper (the Peoria [Illinois] Journal-Star). Because their reports are proprietary and are therefore not accessible online, I know only what I recall from the client’s summary. But the gist of it was that they measured (a) how far a person read before quitting a story and moving on to another and (b) how much of the content they could recall. These are fairly standard measures of readability, similar to the measures Colin Wheildon used in his work in Australia on typography parameters.

Anyway, the results in Peoria—a different time, to be sure, and that’s why this needs to be tested again—suggested that results were quite disparate between men and women. Men were relatively insensitive to which way numbers were presented. They would encounter a number expressed as numerals and keep reading, with similar comprehension either way. Women responded significantly differently, tending to stop reading when they encountered figures and remembering less of the gist of the story if they did read past the figures.

Now we’ve had half a century of consciousness raising about sexism and enrolling girls in STEM and all that, but I’m not sure the center of gravity with respect to societal attitudes has changed as much (in the US, at least) as we coastal elites would like to believe it has. And since we write and edit, usually, for as broad an audience as possible, I think we have to take that center of gravity into account. We need to do some good testing to find out how much progress we’ve made in this regard. Are women still more math-averse than men, or has that evened out for society at large?

Separately, I think context matters. Fiction or nonfiction? STEM or humanities? Etc. On that axis, I think it’s still a judgment call.

That’s interesting, Dick. I haven’t done any research but how about this for a completely non-scientific opinion…. what if the reason women, 50 years ago, responded like this is that in those days women were less likely to be working, and less likely to be working in a ‘digits’ environment. So the reason they stumbled over digits was simply that they were less used to seeing them. Maybe they were reading more books and newspapers than work-related materials, where digits might be more commonly used. Today, the workforce has evened out more (if not fully), so this is less likely to be the case. Women are as likely as men to be reading a lot every day in many different sources and devices (smartphones, social media, etc…) so encounter very similar written content to men. So they are just as accustomed to reading numbers/digits as men are.

I’m sort of inclined to think that one could keep using numbers in fiction, but as the world becomes more digital, and fiction is increasingly read on electronic devices, the usability arguments for digits hold up for fiction too. And as for readability, it’s like any change, such as moving to the metric system, or changing currencies, or changing our vocabulary – you’ll get used to it after a while.

Dick Margulis

Intuitively, I tend to agree with everything you say about the change in women’s circumstances and interests in the last 50 years. But that’s because in my social milieu, all those changes obtain. What I’m proposing is that society as a whole doesn’t change as rapidly as we might wish and believe and that therefore testing is necessary to determine the facts.

That said, as a typographer I know that if we’re going to start using digits wholesale in fiction and narrative prose, we have to be thoughtful about font choices or we’re going to wind up with pages that look like they have measles, and that can pull a reader right out of a story. In technical contexts, lining figures work fine and are generally preferable. In narrative contexts, oldstyle figures are a better choice. But so much text is set by graphic designers with no training in typography, that I suspect much ugliness will ensue.

So true about typography! “But so much text is set by graphic designers with no training in typography, that I suspect much ugliness will ensue.” HAHA, don’t get me started!! 🙂

It’s not easy to find hard information based on usability testing focusing on text readability, and I have never found any research that takes socioeconomic demographics into account. A lot of what you find concentrates on ‘visual’ elements of web pages (pictures and buttons, etc.), interactivity or device usability. When it comes to websites we can look at analytics such as engagement (time spent on a page), which at least can help compare pages but it doesn’t get you very far in drawing reliable conclusions. When it comes to the use of digits or numbers, the web seems to be divided between usability experts (digits!) and people citing standard style guides’ rules for writing out numbers. I go with the usability experts.

Dick Margulis

Okay, I see the problem. You’re seeing the world through a screen. I’m a print guy. (And let’s not talk about ebooks, because they’re still crude in terms of typographical nuance.) Screen reading and dead tree reading remain different experiences.

For the most part, the kinds of material I want to read on a screen are not immersive the way a novel should be. I’m reading short items for ideas and facts on a screen, and in that context, figures make sense. The same applies to shorter print items (news articles, chunks in a how-to manual. If I’m engaged in long narrative text on paper, though, I want to be in a one-on-one conversation with the author, and for that I want as few visual distractions in the text as can be managed.

Leave a Reply

All text content copyright 2016-2022 BelEdit Consulting | Company reg. no.: 2.253.650.191 1 | VAT no.: BE0655.942.605 | All rights reserved. Image sources: Pixabay (with CC0 licence) or BelEdit Consulting.