Precision vs. Accuracy Lab

Close together vs. close to the truth — two different things

Part 1 · Interactive Model

The Target

Move the two sliders. Spread controls how scattered the shots are (precision). Bias controls how far the whole cluster sits from the bullseye (accuracy). Watch which of the four cases you get.

Spread (scatter): low

Bias (off-center): low

Part 2 · Predict Before Reveal

Name the Case

Look at the shots (or the readings). Decide which case it is before the reveal, then lock it in.

Score: 0 / 0

Part 3 · Step-by-Step Machines

Three Reasoning Machines

Pick a data set and its true value. The classifier decides precise? accurate? and names the case.

Data set:

Enter up to four readings and the true value. The machine finds the average and the error.

R1: R2: R3: R4:

True value:

Compare two data sets against the same true value. See which is more precise and which is more accurate.

Compare:

Part 4 · Ten-Question Quiz

Prove It

Question 1 of 10 · Score: 0

Word Box

Accuracy: How close a measurement (or the average of measurements) is to the true value.
Precision: How close repeated measurements are to each other — their spread or scatter.
True value: The actual correct value; the bullseye you are aiming for.
Average (mean): The sum of the readings divided by how many there are; the center of the cluster.
Range (spread): The largest reading minus the smallest; a small range means high precision.
Error: How far the average is from the true value; small error means high accuracy.
Systematic error (bias): A consistent shift in one direction; it hurts accuracy but not precision, and averaging cannot remove it.
Outlier: A single reading far from the others; it widens the spread and lowers precision.