Reading path: Coastal water observation →
Repeat the same procedure
A sensor reading is affected by both the water and the measurement process. Small differences may arise from display resolution, sensor response, placement, or moving water. Repeating the measurement helps reveal how much variation occurs even when conditions appear stable. It does not automatically remove error, but it makes the uncertainty visible.
At each visit, keep the sensor type, depth, orientation, and waiting time consistent. Take several readings at the same station, with enough pause for the device to stabilize according to its instructions. Avoid moving to a new patch of water between replicates unless that is part of the method. Record conditions such as waves, wind, and sunlight, which can make placement harder or cause genuine small-scale variation.
Keep the spread, not only an average
Write down each replicate. You can calculate an average to summarize the visit, but also report the minimum and maximum or another suitable measure of spread. If three readings are 13, 13, and 14 degrees on a whole-degree thermometer, the notebook shows a narrow range at that instrument’s resolution. If they are 12, 14, and 16, the variation deserves attention before interpreting a small difference between dates.
Do not report more decimal places than the instrument supports. An average such as 13.3 degrees may be mathematically produced from whole-degree values, but it can imply a precision the thermometer does not provide. Explain the instrument resolution and retain the original readings. If one value differs sharply, check sensor placement, stabilization, and whether conditions changed during the sequence; do not discard it merely because it complicates the summary.
Compare the variation across visits
A later visit may show a difference larger than the usual spread among replicates. That makes a real change more plausible, but it is not proof if the method or conditions changed. Use the same sampling depth and time where possible, and record tide and weather. If readings vary as much within a visit as between visits, the method may need refinement or the location may be inherently variable at the scale you are observing.
For instance, if replicate readings usually differ by one unit but a matched later visit differs by four, the contrast merits follow-up. Take another set under comparable conditions and inspect the instrument according to its instructions. A careful field conclusion might say “the repeated readings were higher than the previous visit, with little overlap in the observed range,” followed by the conditions. This preserves evidence and uncertainty together. Repetition is most useful when it makes the entire procedure reproducible, not when it simply produces more numbers.
Have a correction or a source to suggest? Send a note →