Not Every Number Means the Same Kind of Certainty
Suppose a box contains
\[ 12 \]books.
If the books were counted correctly, the quantity is exactly \(12\).
Now suppose one book is measured to have length
\[ 23.7\text{ cm}. \]That value came from a measuring process. The true length is not being claimed to equal \(23.700000\ldots\) cm with infinite certainty.
The written value is a reported approximation supported by the measuring instrument and procedure.
\[ \boxed{ \text{exact value} \neq \text{measured approximation} } \]Exact Quantities
An exact quantity is known without measurement uncertainty in the context where it is being used.
Exact quantities commonly arise from:
- counting discrete objects,
- definitions,
- mathematical identities.
For example, a triangle has exactly
\[ 3 \]vertices.
There is no measurement uncertainty in the number \(3\).
Counted Quantities
Counts of discrete objects are typically exact when every object has been identified and counted correctly.
If a shelf contains \(27\) books, then
\[ \boxed{27\text{ books}} \]is exact as a count.
We would not normally interpret it as
\[ 27\pm0.5\text{ books}. \]The count is discrete: there are \(27\) books, not approximately \(27\).
Defined Quantities
Some relationships are exact because the units or concepts are defined that way.
For example,
\[ 1\text{ min}=60\text{ s} \]is an exact conversion relationship in ordinary timekeeping.
Likewise,
\[ 1\text{ m}=100\text{ cm} \]is exact because the centimeter is defined as one hundredth of a meter.
Defined conversion factors do not introduce measurement uncertainty by themselves.
Measured Quantities
A measured quantity is obtained by comparing an unknown quantity with a measurement standard using an instrument or procedure.
Examples include:
- the length of a board,
- the mass of a sample,
- the temperature of a room,
- the elapsed time of a race,
- the voltage across a circuit.
Measured values are not infinitely exact because every measuring process has limitations.
\[ \boxed{ \text{measured value} = \text{best reported estimate} + \text{finite uncertainty} } \]Instrument Resolution
The resolution of an instrument is the smallest change it can reliably distinguish or display.
A ruler marked every millimeter has finer resolution than one marked only every centimeter.
A digital scale displaying to
\[ 0.01\text{ g} \]has a finer displayed resolution than one displaying only to
\[ 1\text{ g}. \]Finer resolution can support more detailed reporting, but resolution alone does not guarantee that the measurement is accurate.
A Reported Measurement Represents a Range
Reported Digits
The digits written in a measurement communicate the scale at which the value has been reported.
For example,
\[ 5.2\text{ cm} \]and
\[ 5.20\text{ cm} \]have the same numerical value as real numbers:
\[ 5.2=5.20. \]But in measurement reporting, the extra zero may indicate that the second value was reported to a finer decimal place.
The notation therefore carries information about measurement reporting, not just numerical equality.
Measurement Uncertainty
Measurement uncertainty describes the range of values reasonably compatible with a measurement and its method.
A measurement might be reported explicitly as
\[ 12.4\pm0.1\text{ cm}. \]This communicates a central reported value and an uncertainty amount.
The notation does not necessarily mean every value in the interval is equally likely. It simply provides a compact uncertainty statement whose precise interpretation depends on context.
The corresponding interval is
\[ 12.3\text{ cm} \leq L \leq 12.5\text{ cm}. \]Measurements as Intervals
A rounded measurement can often be interpreted as representing an interval of possible underlying values.
If a length is reported as
\[ 5.2\text{ cm} \]to the nearest tenth, then values from
\[ 5.15\text{ cm} \]up to but not including
\[ 5.25\text{ cm} \]would ordinarily round to \(5.2\) cm under standard round-to-nearest rules.
One convenient interval notation is
\[ \boxed{ 5.15\leq L<5.25 } \]when the rounding convention is known.
This reveals an important idea: a reported measured value may stand for a range rather than one infinitely precise point.
Accuracy and Precision
The words accuracy and precision describe different ideas.
Accuracy
Accuracy concerns how close a measurement is to the quantity's true or accepted value.
Precision
Precision concerns the fineness or repeatability with which values are reported or measured.
A set of measurements can be tightly clustered but systematically wrong. That would be precise but inaccurate.
A measurement can also be close to the true value by chance while having poor repeatability.
Repeated Measurements
Repeating a measurement can reveal variability in the measurement process.
Suppose repeated readings are
\[ 10.2,\quad10.3,\quad10.2,\quad10.4,\quad10.3\text{ cm}. \]The spread shows that the measuring process does not reproduce exactly the same value every time.
Repeated measurements can help us estimate:
- typical value,
- variation,
- repeatability,
- possible random uncertainty.
A later statistical treatment can describe this variability more formally.
Calculations with Measured Quantities
Arithmetic can transform measured values, but it cannot magically create information that was never measured.
Suppose a rectangle is measured as
\[ 4.2\text{ m} \]by
\[ 3.1\text{ m}. \]A calculator gives
\[ 4.2(3.1)=13.02. \]The arithmetic result is exact for the entered decimals, but the original measurements were not infinitely exact.
Therefore reporting
\[ 13.020000000\text{ m}^2 \]would suggest far more measurement certainty than the inputs support.
\[ \boxed{ \text{calculator digits} \neq \text{measurement precision} } \]Exact Factors in Calculations
Exact numbers do not limit precision in the same way measured quantities do.
Suppose \(5\) identical measured lengths are combined:
\[ 5(2.36\text{ m}). \]The number \(5\) is an exact count.
The measured value
\[ 2.36\text{ m} \]is the quantity carrying measurement uncertainty.
Similarly, the exact conversion
\[ 1\text{ m}=100\text{ cm} \]does not reduce the inherent quality of a measurement when units are changed.
Preview: Significant Figures
Significant figures are one common convention for communicating the precision of measured values and for deciding how many digits to report after calculations.
For example,
\[ 12.30 \]contains more reported measurement detail than
\[ 12.3 \]when both are measurement records.
The rules for significant figures, rounding, decimal places, and uncertainty propagation deserve their own careful treatment.
For now, remember the principle:
\[ \boxed{ \text{do not report more certainty than the data support} } \]Practical Applications
Construction
Cut lengths and tolerances depend on the resolution and reliability of measurement tools.
Science
Experimental results require both measured values and an understanding of their uncertainty.
Engineering
Design calculations must distinguish exact specifications from measured properties and manufacturing tolerances.
Manufacturing
Parts may be accepted only when measured dimensions fall within specified intervals.
Data Collection
Sensor values may contain many displayed digits, but the sensor's actual accuracy and resolution determine how meaningful those digits are.
Worked Examples
Example 1: Exact or Measured?
A class contains \(24\) students.
Assuming the students were counted correctly,
\[ \boxed{24} \]is an exact count.
Example 2: Defined Conversion
The relationship
\[ 1\text{ m}=100\text{ cm} \]is exact.
Thus converting
\[ 2.47\text{ m} \]to
\[ 247\text{ cm} \]does not add uncertainty from the conversion factor itself.
Example 3: Measured Length
A ruler gives
\[ 18.6\text{ cm}. \]This should be interpreted as a measured value with finite precision, not an infinitely exact mathematical constant.
Example 4: Explicit Uncertainty
A mass is reported as
\[ 50.0\pm0.2\text{ g}. \]A simple interval interpretation is
\[ \boxed{ 49.8\text{ g}\leq m\leq50.2\text{ g} }. \]Example 5: Rounded Measurement Interval
A value is reported as
\[ 7.4\text{ s} \]to the nearest tenth.
The underlying value would ordinarily satisfy
\[ \boxed{ 7.35\leq t<7.45 } \]seconds.
Example 6: Calculator Output versus Reported Precision
Measured dimensions are
\[ 2.4\text{ m} \]and
\[ 1.7\text{ m}. \]The calculator product is
\[ 2.4(1.7)=4.08. \]The value \(4.08\) is the exact arithmetic product of the entered decimals, but the measurements themselves do not justify treating all calculator digits as equally certain.
Common Errors
1. Treating Every Written Number as Exact
Measured values carry finite uncertainty even when written as ordinary decimals.
2. Treating Every Integer as Exact
An integer such as
\[ 100\text{ m} \]may still be a rounded measurement. Context determines whether a number is exact.
3. Assuming More Decimal Places Always Mean More Accuracy
Extra displayed digits do not guarantee closeness to the true value.
4. Confusing Precision with Accuracy
A precise measurement can still be systematically wrong.
5. Reporting Every Calculator Digit
Calculation output can contain more digits than the measurements justify.
6. Treating Defined Conversion Factors as Measured Approximations
Exact unit definitions do not introduce uncertainty merely because they are used in a calculation.
Practice
Exact or Measured?
A carton contains \(18\) eggs. Is \(18\) exact or measured?
Check Answer
Assuming the eggs were counted correctly,
\[ \boxed{\text{exact}} \]because it is a discrete count.
Defined Relationship
Is
\[ 1\text{ km}=1000\text{ m} \]exact or measured?
Check Answer
\[ \boxed{\text{exact}} \]because the metric relationship is defined.
Measurement
A thermometer reports
\[ 21.6^\circ\text{C}. \]Is this ordinarily exact?
Check Answer
No. It is a measured value with finite precision.
Interval from Rounding
A length is reported as \(4.8\) cm to the nearest tenth. What interval of values ordinarily rounds to \(4.8\) cm?
Check Answer
\[ \boxed{ 4.75\leq L<4.85 } \]centimeters.
Precision and Accuracy
Can a set of measurements be very close to one another but far from the true value?
Check Answer
Yes. The measurements can be precise but inaccurate.
The Deeper Idea
Measurement turns a physical quantity into a numerical representation.
That representation is useful, but it is not infinitely detailed.
\[ \boxed{ \text{physical quantity} \longrightarrow \text{measurement process} \longrightarrow \text{reported value} } \]This connects measurement to earlier representation ideas:
- the object is not the same thing as its representation,
- a representation may preserve some information while omitting other information,
- scale and units affect how quantities are expressed,
- reported precision tells us something about how finely the representation should be interpreted.
Exact quantities and measured quantities therefore require different habits of reasoning. Exact arithmetic can support unlimited symbolic precision; measured data cannot support more information than the measurement process supplied.
Summary
- Exact quantities have no measurement uncertainty in the context where they are used.
- Counts of discrete objects are commonly exact.
- Defined unit relationships such as \[ 1\text{ m}=100\text{ cm} \] are exact.
- Measured quantities are obtained through instruments or measurement procedures and have finite precision.
- Instrument resolution limits the fineness of distinguishable or displayed changes.
- Reported digits communicate measurement detail, not just numerical value.
- Measurement uncertainty describes a range of plausible values associated with the measurement.
- A rounded measured value can often be interpreted as representing an interval.
- Accuracy concerns closeness to the true or accepted value; precision concerns fineness or repeatability.
- Repeated measurements reveal variability in the measurement process.
- Calculator output can contain more digits than measured inputs justify.
- Exact counts and defined conversion factors do not themselves limit measurement precision.
- Significant figures provide one convention for reporting precision and will be treated more fully later.
The central distinction is simple but powerful: an exact value is known without measurement uncertainty, while a measured value is a finite-resolution estimate of an underlying quantity. Good quantitative reasoning keeps that difference visible throughout calculation and reporting.
Measurement, Scale, and Precision
- Previous: Scale and Reference Quantities
- Measurement, Scale, and Precision — return to the lesson directory.