Goal
Change bias and scatter . Watch which one changes accuracy and which one changes precision.
Accepted value = 100.0 mm
Cause the pattern Explore mode
Systematic bias +0.0 mm
Moves the centre of all readings away from the accepted value.
Random scatter (SD) 0.3 mm
Controls how closely repeated readings agree with one another.
Accurate + Precise
Accurate + Imprecise
Inaccurate + Precise
Inaccurate + Imprecise
↻ New repeats
Reset
Repeated measurements
Bullseye
Number line
Bullseye visualization of repeated measurements
Dots show repeated measurements. The centre is the accepted value. A square shows the mean.
Number line visualization of repeated measurements
Readings are plotted around the accepted value of 100.0 millimetres. The mean is marked separately.
What the pattern means Model thresholds*
Live classification
Accurate & Precise
Mean error 0.00 mm
Spread (SD) 0.30 mm
Precision
readings tightly clustered
Notice: Low bias gives high accuracy. Low random scatter gives high precision.
Progressive practice
Classify the measurement set
Is this set accurate, precise, both, or neither?
Accurate & Precise
Accurate & Imprecise
Inaccurate & Precise
Inaccurate & Imprecise
Use the centre of the cluster for accuracy and its spread for precision.
Show hint
Next challenge
Learning analytics — interaction + quiz trail
0 events ▾
*For this learning model only: “accurate” means |mean error| ≤ 0.5 mm and “precise” means SD ≤ 0.6 mm. In real experimental work, acceptable tolerances depend on the method, instrument and purpose.
×
Two questions to ask
Accuracy: Is the centre/mean close to the accepted value?
Precision: Are repeated readings close to one another?
Systematic error mainly shifts the centre. Random error mainly increases spread.