The measurement chain
Define the variable, choose a suitable instrument and unit, check or zero the tool, use it consistently, then record every value with enough context. Equipment recognition alone is not scientific measurement.
Choose equipment for purpose
Use rulers for length, balances for mass, measuring cylinders for liquid volume, thermometers for temperature and stopwatches for time. Consider the expected range and the size of change, not simply the instrument’s size.
Read analogue scales correctly
Find the smallest division, keep the instrument correctly positioned and view the mark or meniscus at eye level. Defined endpoints and a shared reading rule reduce observer variation.
Use digital tools thoughtfully
Probes and data loggers can collect frequent, consistent readings. They still require correct placement, units, checks or calibration and documented sampling intervals; digital does not mean exact.
Reasonable precision
Record only detail supported by the equipment and method. Extra decimal places do not improve a measurement. Select resolution fine enough to reveal the change relevant to the question.
Resolution is not precision
Resolution is the smallest increment displayed or detected. Repeat-measurement precision is the closeness of repeated readings. A fine-resolution sensor can produce widely scattered, imprecise readings.
Precision is not accuracy
Precise readings cluster closely; accurate readings are close to a trusted value. A miscalibrated instrument may repeatedly produce precise but inaccurate data.
Record usable data
Table headings name variables and units. Keep units, decimal conventions, timing, measurement endpoints and sample labels consistent. Raw observations should remain traceable to their trial or site.
Field observations
Authentic field science may combine site labels, photographs, counts and sensor readings such as soil pH or compaction. Use the same sampling rule and obey safety and permission conditions.
Evaluate limitations
Check unusual readings rather than hiding them. Identify limitations from scale resolution, reaction time, parallax, calibration, positioning or inconsistent sampling, and propose a specific improvement.