Understanding the quality of measurements is fundamental to evaluating evidence and communicating limitations in scientific investigations. VCE Environmental Science requires clear application of these five concepts.
Definition: How close a measured value is to the true (actual) value.
Improving accuracy:
- Calibrate instruments before use (compare against a known standard)
- Repeat measurements and check for consistency
- Use appropriate instruments with sufficient resolution
Example: Using a pH meter that consistently reads 0.3 units too low gives precise but inaccurate pH values. Calibration corrects this.
Definition: How close repeated measurements of the same quantity are to each other — the spread or consistency of results.
Improving precision:
- Take multiple measurements and average them
- Use instruments with finer resolution (e.g. 0.01 g balance vs. 1 g balance)
- Standardise measurement technique to reduce operator variability
Target analogy:
- High accuracy + high precision = all shots clustered at the bullseye
- High precision + low accuracy = shots clustered together but away from bullseye
- Low precision + high accuracy = shots scattered around the bullseye on average
Definition: The degree to which the same researcher, using the same method, equipment and conditions, gets consistent results on repeated measurements.
Example: A student counts bird species at the same point three times within one morning and gets 12, 11 and 12 species — high repeatability.
Why it matters: If results vary greatly when you repeat the same measurement, something is wrong with your method or instrument.
Definition: The degree to which different researchers, using the same method (but possibly different equipment or at different times/locations), can obtain consistent results.
Example: A published vegetation survey method is reproducible if another researcher at a different site, following the same protocol, gets species diversity measures consistent with the patterns reported.
Why it matters: Scientific claims are only considered valid when other researchers can reproduce results — this is why detailed methods must be reported.
Definition: Whether the investigation actually measures what it claims to measure — whether the data collected genuinely reflects the variable being studied.
Example: Using tree height as a measure of ‘ecosystem health’ may not be valid — a forest of tall trees could be highly degraded if understory diversity is absent.
Improving validity:
- Use appropriate indicators that are directly relevant to the research question
- Control extraneous variables that could confound the relationship
- Ensure sampling design captures the full range of variation being studied
| Concept | What It Describes | Main Threat |
|---|---|---|
| Accuracy | Closeness to truth | Systematic error (calibration) |
| Precision | Consistency of repeated measurements | Random error |
| Repeatability | Same method, same researcher, same conditions | Within-experiment variability |
| Reproducibility | Same method, different researchers/settings | Between-researcher and between-site variability |
| Validity | Measuring what you claim to measure | Confounding variables; inappropriate proxies |
When reporting limitations of an investigation:
EXAM TIP: VCAA often asks students to ‘explain how the accuracy/precision/validity of this investigation could be improved’. Always identify the specific source of error first, then suggest a targeted improvement. Vague answers (‘take more measurements’) are insufficient — specify what measurement, why, and how it addresses the identified issue.
