Accuracy, uncertainty and precision
Accuracy, Uncertainty, and Precision
One of the most important goals of science is to produce measurements that accurately represent reality. However, every measurement has limitations. No measuring instrument is perfect, no observer is flawless, and no experiment can eliminate every source of variation. Because of these unavoidable limitations, scientists must evaluate not only the value of a measurement, but also its quality.
Three closely related concepts are used to assess measurement quality:
Accuracy
Precision
Uncertainty
Although these terms are often used interchangeably in everyday conversation, they have very specific and distinct meanings in science. Understanding the differences between them is essential for interpreting experimental data, evaluating scientific claims, and communicating results honestly.
Throughout your chemistry course, you will constantly encounter measurements involving mass, volume, temperature, pressure, concentration, and time. Understanding how scientists evaluate these measurements is just as important as learning how to perform the calculations themselves.
Figure 1: Measuring the Same Quantity Repeatedly
Students measure a metal cylinder's length to understand accuracy, precision, and uncertainty in scientific measurement, highlighting the importance of repeated measurements.
Why Perfect Measurements Do Not Exist
Imagine attempting to measure the length of a pencil using a ruler.
Even if the ruler is well made, several questions immediately arise:
Is the pencil perfectly aligned?
Are your eyes positioned correctly?
Is the ruler itself perfectly calibrated?
Can you estimate between markings consistently?
Each of these factors introduces uncertainty.
Now imagine measuring the mass of a sample on a balance.
Even if the balance displays many digits, those digits do not necessarily represent perfect knowledge. Environmental vibrations, air currents, temperature fluctuations, and instrument limitations all contribute to uncertainty.
For this reason, scientists do not simply report measurements.
They evaluate how trustworthy those measurements are.
Figure 2: Sources of Measurement Variation
Explore various sources of measurement variation in laboratory instruments like balances, rulers, thermometers, and graduated cylinders, highlighting factors such as precision, parallax error, and calibration that affect accuracy.
Accuracy
Accuracy describes how close a measurement is to the true or accepted value.
Accuracy answers the question:
"How correct is the measurement?"
The closer a result is to the accepted value, the more accurate it is.
Example: Measuring a Known Mass
Suppose a calibration standard has a certified mass of:
10.00 g
A student measures the mass and obtains:
10.01 g
This result is extremely close to the accepted value.
Therefore:
The measurement is highly accurate.
Now suppose another student obtains:
9.50 g
This measurement differs substantially from the true value.
Therefore:
The measurement is less accurate.
Figure 3: Accurate versus Inaccurate Measurements
This chart explains the concepts of accuracy and precision in measurements, illustrating how results can be close to the true value (accurate) or consistent with each other (precise) through target diagrams and examples.
Accuracy Is About Correctness
It is important to recognize that accuracy is not determined by the number of digits reported.
For example:
10.000000 g
is not automatically more accurate than:
10.0 g
The extra digits only indicate greater reported precision.
A measurement can contain many digits and still be wrong.
Accuracy depends solely on closeness to the accepted value.
Precision
Precision describes how closely repeated measurements agree with one another.
Precision answers the question:
"How consistent are the measurements?"
Highly precise measurements produce nearly identical results when repeated.
Example: Repeated Measurements
Suppose five measurements of the same sample produce:
9.48 g
9.49 g
9.48 g
9.49 g
9.48 g
These values are extremely close together.
The measurements are therefore:
Highly precise
However, if the true mass is:
10.00 g
the measurements are not accurate.
They are consistently wrong.
Figure 4: Highly Precise Measurements
Illustrating the concept of precision in measurements, this image shows data points closely grouped together, indicating high precision but low accuracy relative to the true value.
Precision Is About Reproducibility
Precision reflects how reproducible a measurement is.
If multiple measurements give nearly identical results, the measurement process is precise.
Scientists often value precision because it suggests the experiment is being performed consistently.
However, precision alone does not guarantee correctness.
You can be consistently wrong.
Accuracy versus Precision
The distinction between accuracy and precision is one of the most important ideas in measurement science.
Students frequently assume that precise measurements are automatically accurate.
They are not.
Accuracy and precision measure completely different things.
The Target Analogy
The classic way to visualize accuracy and precision is with a target.
In this analogy:
The bullseye represents the true value.
Each dart represents a measurement.
Figure 5: Accuracy and Precision Target Diagram
This diagram explains accuracy and precision, demonstrating how measurements can be close to the true value (accuracy) or to each other (precision), with examples ranging from high accuracy and precision to low in both.
High Accuracy and High Precision
In the ideal situation:
Measurements cluster tightly together.
Measurements cluster around the bullseye.
Results are:
Correct
Consistent
This is the goal of scientific measurement.
Interpretation
The measuring device is functioning properly.
The experimental technique is consistent.
The results are trustworthy.
High Precision but Low Accuracy
Suppose the darts form a tight cluster, but the cluster is far from the bullseye.
The measurements are:
Consistent
Reproducible
but
Incorrect
This often indicates a calibration problem.
Interpretation
Something systematic is shifting every measurement away from the true value.
High Accuracy but Low Precision
Suppose the darts scatter around the bullseye.
The average position is correct.
However, individual measurements vary considerably.
The measurements are:
Accurate on average
Not very precise
Interpretation
Random variation is affecting the measurements.
Low Accuracy and Low Precision
Suppose the darts are scattered widely and far from the bullseye.
The measurements are:
Inconsistent
Incorrect
This is the worst-case scenario.
Interpretation
There are likely multiple sources of error affecting the experiment.
Uncertainty
Every measurement contains uncertainty.
This is not a flaw.
It is an unavoidable reality of measurement.
Uncertainty represents the range within which the true value is believed to lie.
Rather than pretending measurements are perfect, scientists explicitly acknowledge uncertainty.
This honesty is one of the strengths of science.
Sources of Uncertainty
Uncertainty can arise from many sources:
Instrument Limitations
No instrument measures perfectly.
A balance reading to 0.01 g cannot reliably determine mass to 0.000001 g.
Human Estimation
Observers must often estimate between markings.
Different people may estimate slightly differently.
Environmental Factors
Temperature fluctuations, vibrations, humidity, and air currents can influence measurements.
Natural Variation
Some quantities naturally fluctuate over time.
Figure 6: Sources of Measurement Uncertainty
Explore the various factors contributing to measurement uncertainty, including instrument resolution, calibration, parallax error, and environmental influences, to improve accuracy in scientific results.
Significant Figures and Uncertainty
Significant figures provide a practical way to communicate uncertainty.
The final reported digit in a measurement is always uncertain.
For example:
4.52 g
implies uncertainty in the hundredths place.
Approximate uncertainty:
±0.01 g
By contrast:
4.5 g
implies uncertainty in the tenths place.
Approximate uncertainty:
±0.1 g
Therefore:
4.52 g is more precise than 4.5 g.
Figure 7: Significant Figures Indicate Precision
Illustrating the concept of significant figures, this chart shows how increasing the number of significant figures in a measurement enhances precision by reducing uncertainty.
Absolute Uncertainty
Uncertainty can be expressed quantitatively.
The first method is:
Absolute Uncertainty
Absolute uncertainty is reported in the same units as the measurement.
Example
4.52 ± 0.01 g
The uncertainty is:
0.01 g
The measurement suggests the true value likely falls between:
4.51 g and 4.53 g
Figure 8: Absolute Uncertainty Range
Understanding absolute uncertainty helps estimate the range within which a true value lies, accounting for potential measurement errors. This concept is crucial for precision in scientific experiments and data analysis.
Relative Uncertainty
Sometimes it is useful to express uncertainty as a percentage.
This allows comparison between measurements of different sizes.
This quantity is called:
Relative Uncertainty
Example
Measurement:
4.52 ± 0.01 g
Calculation:
0.01 g4.52 g=0.0022=0.22%\frac{0.01\,g}{4.52\,g}=0.0022=0.22\%
Relative uncertainty:
0.22%
Why Relative Uncertainty Is Useful
Compare:
100.00 ± 0.01 g
and
1.00 ± 0.01 g
Both have identical absolute uncertainty.
However:
The first measurement is extremely precise.
The second measurement is much less precise.
Relative uncertainty reveals this difference immediately.
Figure 9: Comparing Relative Uncertainty
Comparison of two measurements with identical absolute uncertainty (±0.10), highlighting different relative uncertainties: 10% in Measurement A and 0.10% in Measurement B, emphasizing the significance of relative uncertainty in assessing measurement accuracy.
Sources of Error
Errors affecting measurements generally fall into two categories:
Systematic Error
Random Error
Understanding the difference is essential for diagnosing experimental problems.
Systematic Error
A systematic error is a consistent bias that shifts measurements in one direction.
Systematic errors primarily affect:
Accuracy
Example: Miscalibrated Balance
Suppose a balance always reads:
0.50 g too high
A true mass of:
10.00 g
would be reported as:
10.50 g
every time.
Measurements might be extremely precise, but they are not accurate.
Figure 10: Systematic Error
Illustration of systematic error, demonstrating how consistent bias in measurements can result in values being either consistently higher or lower than the true value, affecting accuracy but not precision.
Common Sources of Systematic Error
Improper calibration
Faulty equipment
Incorrect procedure
Consistent observer bias
Contaminated standards
Why Averaging Does Not Help
Averaging cannot remove systematic error.
Every measurement is biased in the same direction.
The average remains biased.
The only solution is to identify and correct the source of the error.
Random Error
A random error causes measurements to fluctuate unpredictably.
Random errors primarily affect:
Precision
Examples of Random Error
Small temperature changes
Air currents
Vibrations
Reading uncertainty
Instrument noise
These factors cause measurements to scatter above and below the true value.
Figure 11: Random Error
Explore the concept of random error, highlighting how it causes measurement scatter around the true value, affecting precision but not accuracy, and discover methods to mitigate its impact through averaging and multiple measurements.
Why Averaging Helps
Unlike systematic error, random error can often be reduced through repetition.
When many measurements are averaged:
Random highs tend to cancel random lows.
The average becomes more reliable.
This is one reason scientists frequently perform multiple trials.
Uncertainty in Calculations
Whenever measured values are used in calculations, their uncertainties propagate into the final answer.
This is why significant-figure rules exist.
They provide a practical shorthand for tracking uncertainty without requiring advanced statistical analysis.
Example
Calculate:
2.3 × 6.75
Calculator result:
15.525
However:
2.3 has 2 significant figures.
6.75 has 3 significant figures.
The answer must contain:
2 significant figures.
Final answer:
16
Reporting:
15.525
would falsely imply certainty that does not exist.
Figure 12: Propagation of Uncertainty
This infographic explains how uncertainty in measured quantities impacts calculations and limits the precision of final results, using the example of calculating the area of a rectangle.
Scientific Honesty and Reporting Data
One of the most important responsibilities of a scientist is honest reporting.
Scientists must avoid:
Reporting more digits than measurements justify.
Ignoring known uncertainty.
Hiding systematic errors.
Overstating confidence in results.
Doing so creates a misleading picture of the quality of the data.
Scientific integrity requires reporting measurements exactly as accurately and precisely as the data support—no more and no less.
Figure 13: Ethical Scientific Reporting
Ethical scientific reporting emphasizes accurate, honest, and transparent data presentation, highlighting the importance of including uncertainty and using appropriate significant figures to uphold scientific integrity.
Summary of Key Measurement Concepts
Table displaying key terms and definitions related to measurement accuracy, precision, and uncertainties, along with methods for improvement and error identification.
Term | Definition | Improved By |
|---|---|---|
Accuracy | Closeness to the true value | Calibration and elimination of systematic error |
Precision | Reproducibility of measurements | Better technique and repeated measurements |
Absolute Uncertainty | ± range in original units | More precise instruments |
Relative Uncertainty | Percentage uncertainty | Improved precision |
Systematic Error | Consistent bias | Identifying and correcting the source |
Random Error | Unpredictable variation | Averaging repeated measurements |
Figure 14: Summary of Measurement Quality
This infographic explains the elements of measurement quality, highlighting the importance of accuracy, precision, and understanding uncertainty and error to ensure reliable data.
Summary
Accuracy, precision, and uncertainty are fundamental concepts in scientific measurement. Accuracy describes how close a measurement is to the true value, while precision describes how consistently measurements can be reproduced. Every measurement contains uncertainty, and scientists must communicate that uncertainty honestly through appropriate use of significant figures and uncertainty estimates.
Systematic errors reduce accuracy, while random errors reduce precision. Understanding the difference allows scientists to improve measurement quality and evaluate experimental results more effectively. Ultimately, good science requires not only obtaining measurements, but also understanding and communicating their limitations.
Key Points
Accuracy measures closeness to the true value.
Precision measures reproducibility.
Measurements can be accurate, precise, both, or neither.
Every measurement contains uncertainty.
Significant figures communicate uncertainty.
Absolute uncertainty uses the original units.
Relative uncertainty is expressed as a percentage.
Systematic errors affect accuracy.
Random errors affect precision.
Averaging reduces random error but not systematic error.
Significant-figure rules help track uncertainty in calculations.
Honest reporting requires appropriate precision and uncertainty.
Scientific integrity depends on accurately communicating measurement quality.