Math. It’s the universal language: the only form of communication so absolute that it transcends any linguistic and cultural barrier. So definitive that experts in the search for extraterrestrial intelligence (SETI) have chosen math as their preferred language to blast their messages into the stars. While this is a nice idea, the problem with math as the universal language becomes pretty clear when you reach the most critical step: reading. While the principals of mathematics may transcend all the natural languages here on earth, what is not so universal are the numbering systems we use to interpret it.
Time, Space, and Analysis
There are some big differences between how numbers are punctuated in Europe verses the Americas. Our Data Analyst Consultant, Stessie Fertier, who is French, is very familiar with these nuances in numerical representations. Sometimes, for example, spaces are used in place of commas, and sometimes spaces crop up in places where some cultures wouldn’t expect them to—for example, representing one hundred percent with a space between the symbol and the number (100 %) or without (100%).
Time is also represented differently. In France, an “h” is often used to signal hours, for example 18h00, and France usually uses the 24 hour clock. Likewise, the order of day, month, and year varies between countries. Pricing, too, has its differences, as the currency symbol is usually placed after the number ($400 vs 400$). When it comes to data analysis, if the data analyst is not aware of these unique features for their intended audience, it could end up raising an eyebrow at best or at worse the amount in the analysis is off by 100!
“For people who don’t usually think through these kinds of problems, it could be disturbing to them.”
International Students and Decimals
Comma and period placement is also different in many European countries. Mackenzie Mallett holds an International Business degree, during which he had the opportunity to work with Polish exchange students. Eventually they realized through their group work that there was a small difference in numerical representation that could have serious consequences: The Polish students used periods in the place of commas, for example 100.000,00 instead of 100,000.00. While in small classes a simple explanation to a prof would likely suffice, larger classes wouldn’t have the same luxury. In addition, the rise in marking technology would likely complicate things, as the optical character recognition software would not be flexible.
"It was definitely something we had to make them aware of, especially submitting work to Canadian profs. Because if they used any kind of software to mark our work, half our team could fail the assignment."
A Matter of Cost
Our design specialist, Adriana Arjona, also has personal experience with numerical differences: In Colombia, numbering is mostly the same as Canada and other American countries, however, crucial differences lie in their pricing systems. In Colombia, pesos (COP) are the official currency, but like Canada and the USA, uses the dollar sign ($). In addition, because of inflation, COP numbers are typically quite high, which is probably why, similarly to European numbers, for pricing the commas and periods are reversed. For example, $100.000,00 COP is worth around $34.66 CAD and $26.70 USD.
It makes sense why the period is moved forward given these values—most prices would be past the thousand mark—but for those who use both COP and other currencies, it may sometimes be jarring so switch between the two. Online shopping can especially be difficult if the currency is marked only with a dollar sign, as there is a significant difference between a thousand USD and a thousand COP. In addition, in everyday speech, prices over a thousand pesos are shortened to a “mil,” from the Spanish word for a thousand, which English speakers could confuse with a million.
“While in Colombia, I went to a nice restaurant, and when I saw that the bill was $44.000,00, I panicked for a moment until I remembered that the price was in pesos, not dollars, and was worth around $10.50 USD—A very reasonable price.”
Overcoming Biases in Programs and Analysis
Our developer, Colin has noticed differences between French and English number conventions even within Canada, usually in using commas or spaces for the thousand and million marks (1,000 vs 1 000). In addition, the way that individuals write numbers can be problematic for optical character recognition systems, especially in banking software that recognizes cheques.
More broadly, computational programs will always have some kind of bias. From the perspective of many analysts, their goal is to draw conclusions from the data, and sometimes that means smaller items get lost or discounted. If a number is written differently, or if the text is written in another language, it may be dropped from the analysis entirely. Because of this bias, sometimes it’s important to stop and acknowledge that of course, differences in number usage, or other linguistic differences, are not wrong. Even within the same language, many programs are trained on Wikipedia or another source that will have its own biases. If a comment gets dropped, it could be bad for the analysis.
This also relates to length of comments—Some people have a lot to say, while others may only have a couple comments, but that doesn’t mean one is better than the other. It’s important to know your audience and try to adjust your programming ahead of time. Some things are hard to capture in numbers but that doesn’t make them less important.
Overcoming programming bias is something that is important to us in Kai Analytics, especially when it comes to our qualitative analysis platform, Unigrams. Although no analysis can represent every detail, we try find a balance by not losing sight of the bigger picture.
“Discounting different languages could mean improving a product at the expense of others.”
Understanding differences
For many of us, numerical representation was something that we learned when we were very young, and those standards may have been strongly enforced through years of schooling and academia. This can cause us to ignore other conventions elsewhere in the world, and when confronted with these differences we may instinctively feel that our conventions are more “correct.” The reality is that each standard has its own history and reasons behind it, and no choice is more correct than another, just different. It may be most helpful to think of which way is the most helpful for the situation at hand, rather than which way is superior.
One example of this contextual impact is the different ways that numbers are written in type verses handwriting. In Europe, it is common to write 1 with a serif at the top but not at the base. The Arabic numbers we use were originally each meant to represent a number of angles, so the 1 without a base is the older form, but a 1 with a base makes the number more distinct, and the 1 that is just a plain line is doubtlessly the fastest to write by hand. Each method has its own purpose and history, and we can learn to recognize and value each difference.