Reading a result table without over-reading it

A result table looks authoritative.

Rows. Columns. Numbers. Units. Maybe a green checkmark or the word PASS.

That structure can make the information feel more certain than it actually is.

The table is useful. Often extremely useful.

But the table is the output of a measurement process. To read it well, you need to understand what was measured, how it was measured, how the result is expressed, and what the method can reasonably support.

The number is important.

The context around the number is what gives it meaning.

Start with the name of the thing being measured

Before looking at the result, read the row label.

What property is actually being reported?

Depending on the report, a row might describe:

• identity
• assay
• purity
• concentration
• water content
• residual solvent
• mass
• microbial result
• another defined characteristic

Those words are not interchangeable.

A value reported for one characteristic does not automatically answer a different question.

For example, a purity result does not by itself establish quantity.

A quantity result does not by itself establish identity.

And an identity test may establish that a target material is detected without telling you how much of it is present.

Find the method

A result is produced by some analytical procedure.

That matters because different procedures are designed to answer different questions.

Analytical-method guidance treats characteristics such as accuracy, precision, specificity or selectivity, and range as properties that must be considered in relation to the intended purpose of the procedure.

In other words, a method is not simply “good” in the abstract; it has to be suitable for what it is being asked to measure.

So when a result table names a method, ask:

• What is this method intended to measure?
• Is it qualitative or quantitative?
• Is the reported value a direct measurement or a calculated result?
• Is the method information specific enough to understand what was done?

You do not need to become an analytical chemist to ask those questions.

You only need to resist the temptation to let the number outrun the method.

Read the unit every time

Consider these three results:

5

5 mg

5%

They are not three ways of saying the same thing.

A number without its unit can be almost meaningless.

Even percentages need context.

A percentage might represent a mass fraction, area percentage, recovery, relative abundance, or another defined calculation depending on the method and report.

The symbol % tells you the mathematical form of the result.

It does not, by itself, tell you what the percentage represents.

More decimal places do not automatically mean more certainty

A result of:

98.73

looks more precise than:

98.7

And it may be reported that way for a legitimate reason.

But the number of digits printed on a report does not tell you, by itself, how certain the measurement is.

Measurement science treats uncertainty as information associated with a measurement result that characterizes the dispersion of values that could reasonably be attributed to what was measured.

NIST also notes that a measurement result is fundamentally an estimate of the measurand and that its interpretation is incomplete when relevant uncertainty is ignored.

That does not mean every routine result table must display a full uncertainty budget.

It means you should not mistake formatting precision for measurement certainty.

Five digits on a screen are still five digits produced by a real measurement process.

Look for specifications and acceptance criteria

Some tables contain a separate column for a specification, limit, or acceptance criterion.

For example:

Test	Result	Specification
Example characteristic	98.7%	≥ 95.0%

The result and the specification have different jobs.

The result tells you what was reported from the measurement.

The specification tells you the criterion against which that result is being compared.

A PASS usually means that the reported result satisfied the stated criterion.

It does not mean:

• every possible property was tested
• every possible contaminant was excluded
• the sample is universally “good”
• no uncertainty or limitation exists
• the report answers questions outside the listed tests

Understand <, >, ND, and similar qualifiers

Not every result is reported as a simple number.

You may see expressions such as:

• < 0.10
• > 99
• ND
• Not detected
• < LOD
• < LOQ

These are not decorative symbols.

They change the meaning of the result.

Detection and quantitation are different ideas

The detection limit concerns the lowest amount that can be reliably detected.

The quantitation limit concerns the lowest amount that can be quantitatively determined with suitable performance.

ICH Q2(R2), for example, distinguishes the detection limit from the quantitation limit and describes the latter in terms of quantitative determination with suitable precision and accuracy.

NIST likewise defines detection and quantitation limits as distinct concepts.

So:

not detected

does not necessarily mean:

absolutely zero exists

And:

below the quantitation limit

does not necessarily mean:

nothing was detected

It may mean the method cannot support a reliable numerical quantity at that level.

That distinction matters.

Do not turn “not detected” into “absent”

This deserves its own section because it is one of the easiest mistakes to make.

Imagine a method can reliably detect a substance only above a certain level.

If the report says:

Not detected

the defensible interpretation is tied to the capability of that method under those test conditions.

It is not automatically a philosophical statement that zero molecules exist anywhere in the sample.

The method has a boundary.

Good reading respects it.

A method’s range matters

Quantitative procedures operate over a range in which their performance has been established.

ICH Q2(R2) specifically treats range and lower-range limits as part of analytical procedure validation.

It also discusses how performance can change across the calibration range rather than assuming one level of performance everywhere.

That gives you another useful reading habit.

When a result is very near the lower or upper end of a method’s stated range, ask whether the report provides enough context to understand that measurement.

A number being printable does not automatically mean the method has equal performance at every possible value.

Replicates tell you something different from a single result

Sometimes a report contains several measurements of the same property.

You might see:

• replicate 1
• replicate 2
• replicate 3
• average
• standard deviation
• relative standard deviation

That additional information can help you understand the spread of repeated measurements.

Precision, in analytical validation terminology, describes the closeness of agreement among a series of measurements under specified conditions and is commonly expressed through quantities such as standard deviation or coefficient of variation.

But precision has its own limitation:

Measurements can agree closely with each other and still be systematically wrong.

A tight cluster tells you something about repeatability or precision.

It does not automatically establish accuracy.

Accuracy and precision are not synonyms

Imagine three measurements:

9.98
9.99
9.98

They are tightly grouped.

That suggests good precision.

But whether they are accurate depends on the value they are supposed to represent and how that relationship has been established.

Now imagine:

9.4
10.1
10.5

The average might happen to land near an expected value, while the individual measurements are widely scattered.

That is a different problem.

This is why analytical guidance treats accuracy and precision as separate performance characteristics.

A result table may show one, both, or neither.

Do not supply the missing one in your head.

A chromatographic percentage needs its definition

Some analytical reports summarize chromatographic results using peak-area percentages.

Those values can be useful within the defined method.

But the phrase area % matters.

It tells you the result came from the relative signal areas under specified chromatographic conditions.

Do not automatically reinterpret an area percentage as a universal mass-balance statement about everything physically present in the sample.

Different substances can respond differently to a detector, and some things may not be detected by the chosen procedure at all.

The result means what the validated or defined analytical procedure supports.

Not more.

Check whether the result belongs to the sample you think it does

A beautiful result table is useless if you cannot connect it to the material you are evaluating.

Look for identifiers such as:

• sample ID
• lot or batch number
• report number
• test date
• client or submitter identifier
• other traceability fields

Then compare them with the surrounding records.

This is the same principle from Foundation 01:

identity first, interpretation second.

A result without a reliable connection to the sample is just a result belonging to something.

Know when a calculated result is being shown

Not every number in a table comes directly from an instrument.

Some values may be:

• averages
• corrected values
• normalized values
• ratios
• conversions
• calculations derived from other measurements

That is not inherently a problem.

Calculations are part of measurement science.

But a calculated value inherits assumptions and inputs from the calculation used to produce it.

If the report explains the calculation, read it.

If the calculation materially affects your interpretation and is not explained, note the limitation instead of inventing the missing method.

What a result table does not prove

Even a technically sound result table does not automatically prove:

• that the submitted sample represents an entire lot
• that the material remained unchanged after testing
• that every relevant property was tested
• that an unlisted contaminant is absent
• that one method answers questions it was not designed to answer
• that a passing specification is the same thing as universal quality
• that another laboratory would necessarily produce an identical number
• that the document belongs to the material in front of you unless the identifiers connect

The table answers the questions it was built to answer.

Your job is not to make it answer more.

Read the table as a set of bounded answers

A useful result table is not weak because it has limits.

All measurements have scope.

A good report helps you see that scope.

Read the test name.

Read the method.

Read the unit.

Notice the limits.

Separate the result from the specification.

Follow the sample identifiers.

Then ask whether the conclusion you are drawing is actually supported by those pieces.

The goal is not skepticism for its own sake.

It is proportion.

A result should carry exactly as much meaning as the measurement behind it can support.

No less.

And no more.