Skip to content
BENIBIX Learn

Validity vs Reliability

People think a measure that gives the same answer each time must be right. It need not be. Reliability means the measure is steady: do it again and you get the same result. Validity means it measures the right thing. A scale that always reads two kilos too heavy is steady, so it is reliable. It is still wrong, so it is not valid.

General🔬 Research

Validity (Research)

Does the study measure what it says it does? And does the result hold?

Validity asks if a study really measures what it claims to. A spelling quiz is not a fair test of how well you read. Internal validity asks if the result is sound. External validity asks if the result holds for other people and places. A study can be very careful and still miss the mark.

Example
A study says it measures joy by counting smiles in photos. But people smile for cameras, so it is a weak measure.
Why it matters
A precise answer to the wrong question is still wrong. Validity checks that the study is aimed at the right thing.
Full entry
General🔬 Research

Reliability

Whether a measure gives the same result each time it is used.

Reliability is whether a measure gives the same result each time. A reliable test gives a close score if you take it twice. Two people scoring the same answers should also agree. A measure must be reliable to be useful. But steady is not the same as right. A measure can be wrong in the same way every time.

Example
Lucy takes a reading test on Monday and again on Friday. She scores 71 and 73, so the test looks reliable.
Why it matters
If a measure jumps around, no single score can be trusted. So it has to come first.
Full entry

In one line

Validity (Research)
Does the study measure what it says it does? And does the result hold?
Reliability
Whether a measure gives the same result each time it is used.