Air quality monitors are now being widely used. But we’ve found that perhaps the most important question is not being answered: how accurate is the data?
While monitoring tech is indeed enabling new insights at lower costs, higher data resolutions, and with less maintenance, these insights are not useful, and perhaps even harmful, if the data isn’t accurate.
Indicative air quality monitors should go through a “co-location” study. The monitor is deployed with a regulatory reference instrument to compare the readings against one another. This will give a user of these monitors a sense for baseline accuracy, but also for any calibration needs.
What’s Calibration? Learn how to set up a co-location for indicative monitoring with Clarity’s free Calibration Guide. Download here.
In our "How to Assess Sensor Accuracy" series, we share how to evaluate the data accuracy for indicative monitors using two of the most commonly employed statistical measures: the Pearson squared correlation coefficient and Mean Absolute Error.
1. The Pearson squared correlation coefficient (R²) is calculated to determine how measurements from a device under analysis correlate with measurements of a reference instrument, or in other words, how well the device under analysis measures changes in pollutant concentration compared to the reference instrument.
Read about how to use R² to assess sensor accuracy here.
2. Mean Absolute Error (MAE) is calculated to determine the absolute difference between the measurements of a device under analysis and the measurements of a reference instrument. It is the average of the absolute value of the deviations between measurements of the two devices. In other words, it shows how different measurements from the two devices are in value.
Read about how to use MAE to assess sensor accuracy here.
Both MAE and R² should be considered together to get a good picture of the accuracy of the device under analysis. Here’s a suggested summary table on how to use the two metrics together: