Wait, what? A sensor display stays at 10.0 while the true input moves from 10.0 to 10.1. A student says, “Nothing changed. The screen proves it.” But the instrument specification contains one quiet line: Dead band: 0.2 unit.
The unchanged display is an observation. “The physical quantity did not change” is an inference. Those are not automatically the same thing.
Quick Answer
A measuring system can sometimes experience a small change in the quantity being measured without showing a detectable change in its indication. The International Vocabulary of Metrology describes a dead band as an interval through which the measured quantity can change in both directions without producing a detectable change in the indication.
So a steady display is evidence that the instrument did not indicate a detectable change. It is not automatically proof that the underlying quantity was perfectly constant.
The Exact Learner Job
This Reality Lab owns one narrow job: evaluating a claim of “no change” when that claim is based on a measuring instrument whose dead band may hide small changes.
It does not replace the canonical PSLE Science owners for reading scales, repeated measurements, precision, fair tests, variables or conclusions. It applies those skills to a real specification-sheet problem: what does an unchanged indication actually allow us to conclude?
Build an Original Case
Consider a fictional force display with a dead band of 0.2 N. A reference device changes the force in tiny steps while the test display is watched.
| Reference force | Displayed force |
|---|---|
| 5.00 N | 5.0 N |
| 5.05 N | 5.0 N |
| 5.10 N | 5.0 N |
| 5.15 N | 5.0 N |
| 5.25 N | 5.2 N |
The display staying at 5.0 N from the first to the fourth row does not prove the force was unchanged. The reference shows that the physical input did change. The test instrument simply did not show a detectable change over those smaller steps.
Observed vs Claimed vs Inferred
| Evidence layer | Statement |
|---|---|
| Observed | The displayed number stayed at 5.0 N for several small input changes. |
| Claimed | The specification states a dead band of 0.2 N. |
| Inferred | Small real changes may occur without producing a detectable change on this display. |
The scientific habit is to stop at the strongest statement the evidence supports. “The indication did not change” is stronger evidence than “nothing in the system changed.”
Do Not Confuse Dead Band With Resolution
Two ideas can look similar but answer different questions. Resolution concerns the smallest change that can be meaningfully distinguished or displayed under stated conditions. Dead band concerns a region in which the input can move without producing a detectable change in indication. A display can show many digits and still have a dead band caused by the behaviour of the measuring system.
That is why the number of decimal places on a screen does not by itself tell you how small a real change the instrument can reliably reveal.
Do Not Confuse Dead Band With Accuracy Either
An instrument may have a small dead band yet still show a biased value. Another may be well calibrated at certain points but have a wider dead band around them. Accuracy, calibration, resolution, repeatability, response and dead band are related parts of measurement performance, not interchangeable labels.
Representation Check: A Flat Line Can Hide Motion
Imagine a graph made from the instrument display. It forms a flat horizontal line for ten seconds. A flat line means the recorded indications did not change. It does not prove the actual input was perfectly flat unless the instrument was capable of revealing all changes that matter to the question.
This is one reason scientific graphs must be read together with their measurement method. The graph is a representation of measured data, not a magical window into every small change in reality.
Method Check: How Would You Test for a Dead Band?
- Use a reference source that can change the input in controlled small steps.
- Record both the reference value and the instrument indication.
- Approach the region from more than one direction when appropriate.
- Allow enough time for the instrument to respond so simple delay is not mistaken for dead band.
- Repeat the sequence to see whether the no-change region is consistent.
- Keep temperature, power supply, mounting and other relevant conditions stable.
The Crucial Alternative Explanations
A display that does not move could have several explanations: the input truly stayed constant; the input changed by less than the instrument can reveal; the instrument has a dead band; the sensor responds slowly; the display rounds the value; the data logger updates only at intervals; or the measurement system has become stuck or saturated.
Good reasoning asks which explanation the method can separate.
Worked Case 1: The Water Tank Controller
A tank level controller displays 50 cm for several minutes while a small amount of water is added. An advertisement claims, “Rock-steady sensing: water level never changes unexpectedly.”
The display alone cannot support that claim. The added water gives an independent reason to expect some change. To decide whether the level change is too small for the instrument to reveal, we need the sensor specification, the tank geometry, and preferably a more sensitive reference measurement.
Worked Case 2: A Thermostat
A thermostat turns a heater on below one temperature and does not switch it off until a higher temperature is reached. The displayed temperature may remain unchanged for small fluctuations. A student sees the fixed display and concludes that the room temperature is perfectly constant.
That conclusion is too strong. Control systems can intentionally include regions in which small changes do not trigger a new action, and measurement systems can also have limited sensitivity to small changes. The learner should separate no displayed change, no control action and no physical change.
Worked Case 3: Two Product Claims
Sensor A advertises “dead band ≤0.05 unit.” Sensor B advertises “dead band ≤0.20 unit.” A comparison page says Sensor A is “four times more accurate.” That does not follow. The figures describe one aspect of detecting change. To compare overall measurement quality, we would need evidence about accuracy, calibration, repeatability, range, environment and intended use.
What Evidence Strengthens the Claim That Small Changes Are Being Hidden?
- An independent reference records small input changes while the test display remains fixed.
- The no-change interval repeats over several trials.
- The display eventually changes after the accumulated input crosses a consistent threshold.
- The behaviour remains after adequate settling time.
- The pattern appears in both directions in a way consistent with the stated dead-band definition.
What Evidence Weakens It?
- The reference also shows no change.
- The apparent dead band disappears when the update rate is increased.
- The sensor was simply still settling after each step.
- The display was rounded while higher-resolution raw data changed continuously.
- The result occurs once and cannot be repeated.
Tempting Reasoning Traps
| Trap | Repair |
|---|---|
| “The number did not move, so the quantity did not move.” | State only that no detectable change appeared on that instrument. |
| “More decimal places remove dead band.” | Displayed digits and physical response are not the same property. |
| “Dead band 0.2 means every reading is wrong by 0.2.” | Dead band describes a no-detectable-change interval, not a fixed error added to all readings. |
| “A smaller dead band proves a better product overall.” | It supports one narrower performance comparison only. |
How Far Can a Conclusion Travel?
A dead-band value measured under one set of conditions does not guarantee identical behaviour at every temperature, speed of change, mounting arrangement or operating range. The BIPM vocabulary itself notes that dead band can depend on rate of change. That detail matters: the speed of the input may be part of the evidence.
PSLE-Style Transfer Case
A digital sensor records 7.0 units for three consecutive readings. A more sensitive reference instrument records 7.00, 7.06 and 7.11 units at the same moments.
A careful explanation is: The unchanged sensor display does not prove the measured quantity stayed constant because the reference recorded small increases. The test sensor may not have been able to indicate changes of that size. The learner should check the sensor’s measurement characteristics and repeat the comparison before deciding why.
Delayed Independent Return
Come back tomorrow and explain, without using the word “dead band” at first, why a constant display does not always prove a constant physical quantity. Then name two alternative explanations for the same observation.
Explained Practice
- A display stays at 2.0 while a reference moves from 2.00 to 2.08. What may you conclude? The test instrument did not display that small change; do not conclude the quantity was constant.
- A display changes only every 0.1 unit. Does that prove a 0.1-unit dead band? No. Rounding or display resolution could also explain the steps.
- Why repeat the test from both directions? Because direction-dependent behaviour can reveal additional measurement effects and helps distinguish patterns.
Route to Existing eduKateSengkang Owners
Use this Reality Lab page as the application layer, then return to repeated-results reasoning, random variation and systematic shift, and the difference between a set condition and the condition actually experienced.
Parent and Tutor Teaching Guide
Give the learner two rulers: one marked every centimetre and one marked every millimetre. Move a pencil by a tiny amount and ask what each ruler can reveal. Then explain that real instruments add more complications than scale markings alone. The core question is, “What size of change could this measurement system actually show?”
Avoid turning the lesson into a vocabulary quiz. The durable habit is to refuse the jump from “I cannot see a change” to “there was no change.”
Authoritative Sources
- BIPM/JCGM International Vocabulary of Metrology: Dead Band — formal definition and note that the effect may depend on rate of change.
- NIST: Noise Averaging and Measurement Resolution — background on digital measurement resolution and quantisation.
- SEAB 2026 PSLE Science syllabus — current assessment objectives include interpreting and analysing information and evaluating observations, information and methods.
The Quiet Habit
When a display does not move, translate that observation carefully: no detectable change appeared here. Then ask whether the measurement system could have hidden a smaller real change. That single question prevents a screen from becoming stronger evidence than it really is.
