Conversion errors and the metric system


A recent headline on a metric system conversion error

I recently received news that the Centers for Disease Control and Prevention (CDC) made a major conversion error relating to the metric system. The upshot is that it greatly underestimated the risk of formaldehyde in laminate flooring. The underlying mistake: it failed to convert between meters and feet initially reporting the estimated risk at one-third of what it should have been.

The organization did not come out and say lack of metric adoption was the cause of the error:

The CDC/ATSDR indoor air model used an incorrect value for ceiling height.  As a result, the health risks were calculated using airborne concentration estimates about 3 times lower than they should have been.

However, others were more than happy to point out the real root of the problem:

CDC fixes major error in flooring risk report: Not converting to metric – Retraction watch

CDC Revises Health Risk Assessment Of Flooring After Math ErrorCDC recently announced that laminate floors are safe, only to realize that they forgot to convert from feet to meters—and that the cancer risk is three-fold higher –

There are some who believe that conversions are easy to make and therefore, living with two measurement systems shouldn’t be a problem.

Marciano's book

Marciano’s book

In fact, in his book, Whatever happened to the metric system: How America kept its feet, John Bemelmans Marciano (Kindle location 2020 for both quotes), states:

Conversion is now as easy as speaking “seven ounces to grams” into your smartphone and immediately receiving the answer 198.446662g.

Marciano later goes on to say:

Why would Americans go metric when computers have done the job for them and they don’t even have to know about it?

How about a three times greater risk for potential negative health effects due to human unit-confusion error?

Luckily, the CDC was able to quickly make a correction but who knows how many other errors haven’t been caught and continue to put us at risk in one way or another?

The idea that technology will save us from conversion errors is flawed because it assumes that the human element won’t impose the error.

Surely the CDC has access to computers and other high-tech gadgets at least as good, if not better, than what I have access to in my smartphone and yet, the mistake was still made.

Again, it’s not a technology issue, it’s a human issue that will always occur even if the frequency of such mistakes is not currently well known.

Aside from outright errors, there’s the time it takes to make a conversion in the first place. Add up the time it takes to whip out the cell phone, ask the question, wait for the answer and read it. Then one needs to multiply that by how many people in this country need to do that in a year. All wasted time. One set of measures eliminates the entire issue.

I’ve previously pointed out people are already at risk every time their pharmacist converts a prescription written in milliliters (as they all are) into teaspoons and tablespoons. Why are we doing this to ourselves?

Conversion errors are inevitable  

While I so far have been unable to find any statistics on how often conversion errors occur, everyone seems to recognize they do happen and research seeks interfaces that try to minimize them. One paper I reviewed, Reducing number entry errors: solving a widespread, serious problem by Thimbleby and Cairns indicates:

Ironically, the more skilled a user, the less attention they will pay to what ought to be routine outcomes, so the more likely these types of error will go unnoticed until they have untoward consequences. The reason is, as users become skilled, they automate actions, so their attention can be used more selectively; thus as they become more skilled, they pay less attention to the display, whose routine behaviour they have learnt to expect (Wickens & Hollands 2000).

A conversion "helper" from the 1970s

A conversion “helper” from the 1970s

Still, we can learn from our past. One of the things I’ve heard from people regarding our last attempt at metric adoption in the 1970s (I was a bit young at the time to remember) was students were taught difficult and confusing conversion formulas.

Next time, just have people start using the new, metric system measures and convert only those things that are absolutely necessary. Fewer conversions means fewer errors.

Thanks for reading.

More next month.