When breath alcohol test results change in an environmental test that alters temperature and humidity together, safety managers and quality or engineering staff need to tell whether the difference reflects a change in the sensor's response or a difference in how the test was prepared. An electrochemical sensor in particular reads the electrochemical reaction to alcohol in the breath as an electrical signal and displays it as a concentration, so interpreting results involves not only the test environment but also how long the device stayed in that environment, the sample conditions, and the order of tests.
For example, the official information for DATECH's AF-40BT gives an operating temperature of -5 to 50 ℃, and gives accuracy under conditions of 25 ℃ and a specific concentration. This refers to the stated accuracy standard under that temperature condition. It does not mean the same result has been confirmed in every environment combining temperature and humidity. So when results differ between two environments, rather than first labeling it a "temperature effect" or "humidity effect," read it by first separating what was changed and what was held fixed.
Results from simultaneous changes compare two mixed factors
For example, if both temperature and humidity were changed between a reference environment and a comparison environment, the difference between the two results cannot be pinned down as the temperature change, the humidity change, or the effect of both conditions changing together. The observation that the value in the comparison environment was higher or lower is itself valid, but that observation alone does not support concluding that a particular environmental factor was the cause.
What matters here is not eliminating the difference in results, but accurately limiting the question the difference can answer. A comparison of two combinations in which temperature and humidity both changed can answer "was there a difference between these two environment combinations?" It cannot answer "how much does it change when only temperature changes?" or "is humidity the cause?"
Build comparison axes around a reference combination
To separate the effects of temperature and humidity, set one combination as the reference and include combinations that change only one variable from that reference. This does not mean recommending particular temperature or humidity values. It means designing the comparison structure within the device's allowed conditions and the purpose of the test. The device model, firmware version, the sameness of the sample or test gas used, and the number and order of tests are fixed items as important as the environmental conditions.
| Comparison combination | Condition changed from the reference | What this comparison can tell you |
|---|---|---|
| Reference environment | Temperature and humidity kept at reference values | Test results under reference conditions |
| Temperature only changed | Humidity kept, only temperature changed | The temperature difference at that humidity |
| Humidity only changed | Temperature kept, only humidity changed | The humidity difference at that temperature |
| Temperature and humidity changed together | Both temperature and humidity changed | The overall difference between the two environment combinations |
The reason to split the combinations as in the table is simple. The difference between the reference environment and the "temperature only changed" combination can be read as a temperature difference only when humidity is the same. The same principle applies to the reference environment and the "humidity only changed" combination. The combination with both temperature and humidity changed serves as data for seeing whether there is a difference that the first two comparisons do not fully explain, but it does not replace evidence for a single factor.
Record observed values together with the scope of interpretation
Take a hypothetical example. Suppose repeated tests were run with the same device and the same test sample in the reference environment, an environment with only temperature changed, an environment with only humidity changed, and an environment with both changed. The observations here can be the displayed value for each combination, the spread of the repeated results, and whether any tests failed or ended abnormally. Even if the average changed in a particular combination, unless you also look at whether the repeated results were consistent, it is hard to separate changes due to the environment from fluctuation in individual tests.
Limit the interpretation to each comparison combination. If a difference was observed in the combination with only temperature changed, write "a difference was observed in the comparison that changed temperature under a fixed humidity condition." Record the combination with only humidity changed the same way. If a larger difference appeared in the combination with both conditions changed, you can consider the possibility of an interaction between temperature and humidity, but you cannot conclude that it exists. An interaction means a case in which the effect of one condition differs depending on another condition.
A possible action is to repeat the combination that showed a difference and compare how long the device stayed in each environment before testing with the actual temperature and humidity records. It is also more appropriate to set the completion criterion not as "the results became the same," but as a state in which you can trace which conditions were held fixed for each comparison and the reasons for handling outliers are recorded. If you excluded results, do not simply delete them. Keep a record, together with the original results, of which tests were left out of the comparison and why.
Test order and preparation can look like environmental effects
In environmental tests, the difference between the combination run first and the one run later can look like an environmental difference. This is because the device adjusting to the environment, differences in sample preparation, and changes in conditions from repeated testing can overlap with the order. So you need to record the start and end times of each combination, how long the device was placed in that environment, and the actual order of execution, so the result differences can be reviewed again.
It is also safer not to treat ad hoc repeats using human breath as having the same meaning as samples for environmental testing. A breath sample can change with the person's condition and the way it is collected, so it becomes an extra variable in a comparison meant to isolate the effects of temperature and humidity. If the purpose of the test is to confirm the device's environmental characteristics, you need to explain how the sameness of the sample was secured before presenting the results table.
Write the report's conclusion only for the combinations confirmed
For a product like the AF-40BT, whose accuracy is stated at 25 ℃ and a defined concentration, you must not transfer that figure as a guaranteed value for other temperature and humidity combinations. The operating temperature information is also guidance on the temperature range in which the device can be used. It is not data that stands in for the measurement characteristics of every combination, including humidity conditions.
In the conclusion, state "which temperature and humidity combinations were compared," "which of the model, sample, and test order were held fixed," and "which combinations have not been tested yet." Leaving the scope this way avoids overgeneralizing results from tests that changed temperature and humidity together, while making clear what needs to be checked separately in the next test. The key to a test that changes temperature and humidity together is not rushing to pick a single cause, but limiting the scope of what the result difference tells you as precisely as the comparison design.
