Follow

# I have two Home devices alongside and they show different values. How is that possible?

It is perfectly possible!

When it comes to Radon measurements, the data to consider is something called Standard Deviation.

The Standard Deviation measures the level of dispersion of a set of values: when we say, for example, that Radon levels are 4 pCi/L (or 150 Bq/m3), we are saying that there is a distribution of data made by a set of values captured, and most of these values are around 4 pCi/L (or 150 Bq/m3). It’s a statistical concept.

Now, let's see a practical example: the standard deviation of a product like the Corentium Home is ±20% at 2.7 pCi/L (or 100 Bq/m3) after 7 days. This implies that the distribution of data can be 20% broader than 2.7 pCi/L, i.e. 0.54 pCi/L (20 Bq) lower and 0.54 pCi/L (20 Bq) higher than the reference value.

In other words, it is hypothetically acceptable that one monitor is measuring at 3.24 pCi/L (or 120 Bq/m3) and one is measuring at 2.16 pCi/L (or 80 Bq/m3), as they are within range. Bear in mind two things:
1. the standard deviation improves during longer periods of time and it falls under 10% after one month.
2. this type of accuracy is very precise when it comes to Radon detection. Radon is a very difficult gas to measure and this standard deviation is considered very good.
In conclusion, not only is this normal and within the statistical error widely accepted for Radon measurement; it also means that in the short-term the values can (and do!) differ greatly. Radon levels fluctuate very much even from room to room, so differences are normal. This is also the reason why we recommend you let your monitors measure Radon for as long as possible, and you will see that the values of the two monitors eventually tend to converge.
10 out of 12 found this helpful
Have more questions? Submit a request

• Russ

I have two Corentium Home that I use side-by-side, and after a year of moving them around the house, they have always been very close and at times, exactly the same in the Long Term value.

• Sten Kaiser

In the sales material and in the manual you say that the error after 7 days is +-10%. Here you say that the standard deviation is 20%. Errors are usually given with 95% confidence and a 20% standard deviation gives an error of +-40% at 95% confidence (and +-50% at 99%).

So is the error +-10% or +-40%? It quite a difference.

The article also uses the standard deviation in a weird way.