Systematic and random errorsWhat kind of error it is, what it does to your results, and how to reduce it
No measurement is perfect. What matters is knowing which kind of error you are dealing with, because each kind affects results differently and each has a different fix. This is one of the most tested ideas in practical science.
Years 7 to 12
Physics, Chemistry, Biology, Science
Updated November 2026
The short answer
Systematic error
Pushes every result the same way.
Every reading is too high, or every reading is too low, by a similar amount. It affects accuracy.
Random error
Makes results scatter unpredictably.
Some readings are a bit high and some a bit low. It affects precision and reliability.
Mistake
A one-off slip.
Misreading a scale, spilling a solution or writing the wrong number. The fix is to redo that measurement.
Side by side
Systematic error
Random error
What it does
Shifts every result in the same direction
Scatters results above and below the true value
What it affects
Accuracy
Precision and reliability
Does repeating help?
No. The same error is in every trial.
Yes. Averaging repeats reduces its effect on the mean.
How to reduce it
Improve the method or the equipment: calibrate, zero, insulate
Repeat and average, improve technique, use finer instruments
On a graph
The whole line shifts, often giving an intercept that shouldn’t be there
Points scatter either side of the line of best fit
Common errors and their fixes
Error
Type
Why
How to reduce it
Balance not zeroed (zero error)
Systematic
Every mass is out by the same amount
Press tare before each use
Wrongly calibrated thermometer
Systematic
Every temperature reads high or low
Check it in a mixture of ice and water, which should read 0 °C
Heat lost in a calorimetry experiment
Systematic
Temperature rise is always smaller than it should be
Lid, insulation, draught shield
Air resistance on a falling object
Systematic
Always slows the object down
Use a dense, compact object or a short drop
Reaction time when timing by hand
Mostly random
The delay is different each time
Time many swings or use light gates
Judging the end point of a titration
Random
The colour change is seen at slightly different points
Add drop by drop near the end point and repeat to concordant titres
A reading that flickers on a meter
Random
The value changes while you read it
Record the middle of the range, or use the mean of several readings
Is parallax random or systematic?
Parallax error happens when you read a scale from an angle instead of straight on, so the pointer or meniscus seems to line up with the wrong mark.
It depends on how it happens. If you read from the same wrong angle every time, such as always looking down at a measuring cylinder on the bench, every reading is off in the same direction. That is a systematic error. If your eye is at a slightly different height each time, the readings scatter, which is a random error.
Either way, the fix is the same: get your eye level with the mark or the bottom of the meniscus.
Why “human error” isn’t an answer
Writing “human error” tells a marker nothing. Every error involves a human somewhere. Name the actual error, say whether it is systematic or random, and say how it affected the result.
Weak
There may have been human error in timing.
Strong
Timing by hand introduced a random error from reaction time, of about 0.2 s at the start and stop. Timing 10 swings instead of one would reduce its effect on each period.
Names the error, its type and size, and a fix.
Weak
The results were wrong because of errors.
Strong
The calculated heat of combustion was 38% lower than the accepted value. This is mainly a systematic error: heat was lost to the air and the metal can, so the water absorbed less energy than the ethanol released.
Gives the size and direction of the error and explains the cause.
A practical, worked through
Heat of combustion of ethanol
Chemistry, Years 11 and 12
A spirit burner heats 200 mL of water in a metal can. The temperature rise of the water and the mass of ethanol burned are used to calculate the heat of combustion. School results usually come out well below the accepted value.
Systematic errors
Heat lost to the air and to the can, so the water absorbs less than the ethanol releases. Incomplete combustion, shown by soot on the can, releases less energy. Ethanol evaporating from the wick counts as “burned” mass. All three make the result too low.
Random errors
Reading the thermometer as the temperature changes, draughts that vary from trial to trial, and the uncertainty in the small mass of ethanol burned.
Improvements
Use a lid and a draught shield, keep the flame close to the can, stir the water, and put the cap on the burner as soon as it is put out. Repeat and average to reduce the random errors.
A metre ruler is used from its end, which is worn down by 2 mm.
Systematic. Every length is out by the same 2 mm.
Students time a trolley rolling down a ramp with a hand-held stopwatch.
Random. Reaction time is different each time, so the times scatter.
A spring balance reads 0.2 N when nothing is hanging from it.
Systematic. A zero error adds 0.2 N to every reading.
The temperature of the room changes slightly during a long experiment.
Random. The effect goes up and down unpredictably across the trials.
A student always reads the measuring cylinder from above.
Systematic. Reading from the same wrong angle puts every reading off in the same direction.
A balance reading drifts up and down because of draughts in the room.
Random. The reading varies unpredictably, so it is random.
The thermometer in a calorimetry experiment is accurate, but heat escapes from the cup.
Systematic. Heat loss always makes the temperature rise too small.
Different students judge the colour change at the end point slightly differently.
Random. The judgement varies from trial to trial in both directions.
Common mistakes
Writing “human error”.
Instead: Name the specific error and its type.
Saying repeating removes all error.
Instead: Repeating reduces the effect of random error only. Systematic error stays.
Saying random error can be eliminated.
Instead: It can be reduced, never removed completely.
Calling a mistake an error.
Instead: A slip like writing down the wrong number is a mistake. Redo that measurement.
Not saying which way the error pushed the result.
Instead: For a systematic error, say whether it made the result too high or too low.
Suggesting a fix that doesn’t match the error.
Instead: Repeat trials for random error. Change the method or equipment for systematic error.
Exam-style questions
Distinguish between systematic and random errors. Give an example of each. 3 marks
Show a model answer
A systematic error shifts every measurement in the same direction by a similar amount, such as a balance that hasn’t been zeroed. A random error causes measurements to scatter unpredictably above and below the true value, such as reaction time when timing by hand. Systematic errors affect accuracy; random errors affect precision.
Explain why repeating a measurement does not reduce a systematic error. 2 marks
Show a model answer
A systematic error affects every trial in the same way and by a similar amount. Each repeat contains the same error, so averaging them gives a mean that is still shifted by that amount.
A student plots extension of a spring against the mass hung on it. The line of best fit is straight but crosses the y-axis at 0.8 cm instead of passing through the origin. Explain what this suggests, and whether it affects the spring constant calculated from the gradient. 3 marks
Show a model answer
It suggests a systematic error, such as a zero error in measuring the starting length, which added about 0.8 cm to every extension. Because every point is shifted by the same amount, the gradient is not affected, so the spring constant calculated from it is still valid.
Identify the type of error caused by always reading a measuring cylinder from above, and state how to avoid it. 2 marks
Show a model answer
Parallax error, which here is systematic because every reading is off in the same direction. Avoid it by reading the bottom of the meniscus with your eye level with it.
Questions students and parents ask
Is parallax error random or systematic?
It can be either. Reading from the same wrong angle every time makes it systematic. Reading from a slightly different angle each time makes it random. Reading at eye level fixes both.
Is reaction time a random or systematic error?
Mostly random, because the delay is different each time you start and stop a stopwatch. Timing many oscillations, or using light gates or video, reduces it.
Is human error a type of error?
Not in science reports. It is too vague. Name the actual error, such as parallax or reaction time, and say whether it is random or systematic.
Can errors be eliminated completely?
No. Errors can be reduced but never removed completely. That is why results are given with an uncertainty.
What is a zero error?
An instrument that doesn’t read zero when it should, such as a balance showing 0.15 g with nothing on it. It is a systematic error that shifts every reading by the same amount.