What this lesson teaches
I can tell accuracy, reliability and validity apart and say how to improve each one.
Syllabus 9173, Fieldwork, 1. Pre-fieldwork stage. Students should be able to develop a plan that: ensures accuracy and reliability of data collected.
Make a guess
A thermometer reads exactly 2 degrees too high every time. Are its readings reliable, accurate, or both?
- Neither, because thermometers cannot be trusted.
- Both, because the error is the same each time.
- Accurate but not reliable.
- Reliable but not accurate.
Show the answer
Reliable but not accurate.
It gives the same result every time, so it is consistent. But every reading is 2 degrees from the true value, so it is not accurate.
Accuracy is closeness to the true value, reliability is consistency, and validity is whether the evidence answers your real question.
Accuracy means how close a measurement is to the true value. A thermometer that always reads 2 degrees too high is consistent, but not accurate. Check instruments against a trusted one first. Showing many decimal places does not fix a wrong reading.
Reliability means getting the same result when you repeat a measurement in the same conditions. Train observers, test questions first and repeat readings. A real change between noon and evening is not unreliability. Compare like with like.
Validity means whether the method measures what you intend and supports your conclusion. A reliable method is not always valid. Counting benches may be reliable, but it does not measure accessibility if the benches are hard to reach. Combine indicators and ask users.
Sort these
Which kind of problem does each fieldwork issue mainly show?
Groups: Accuracy problem, Reliability problem, Validity problem.
- A rain gauge leaks, so it always records less than fell
- Two observers count very different numbers of walkers at the same spot and time
- Asking only about distance when shade and steps may matter more
- Untrained students score the same street differently on a quality scale
- Counting bus stops to judge whether older residents can reach a clinic
- A tape measure has stretched, so every width reads too short
Show the answer
Accuracy problem
- A rain gauge leaks, so it always records less than fell The readings are consistently below the true amount, which is an accuracy problem.
- A tape measure has stretched, so every width reads too short Each reading is off from the true value in the same way, so the problem is accuracy.
Reliability problem
- Two observers count very different numbers of walkers at the same spot and time Repeating the measurement in the same conditions gives different results, so it is not reliable.
- Untrained students score the same street differently on a quality scale Without agreed rules, the same place gets inconsistent scores, which is unreliable.
Validity problem
- Asking only about distance when shade and steps may matter more The method does not fully measure what the question asks, so the conclusion lacks validity.
- Counting bus stops to judge whether older residents can reach a clinic The count may be consistent, but it does not measure whether people can actually make the journey.
Match each fix to its problem. Calibration fixes a biased instrument. Repeat readings reveal random errors. Sampling missing groups improves coverage.
- Accuracy
- Close to the true value.
- Reliability
- Consistent when repeated.
- Validity
- Measures what you actually intend.
Step by step
Find the problem
Instrument, sampling or interpretation error?
Explain its effect
How could it change the result?
Target the fix
Choose a change that tackles that cause.
Worked example: A strong evaluation
"Observers used different ideas of shade, so scores may reflect the observer, not the site. Agree a rule for shade, practise together and compare ratings before the main survey." This links problem, effect and fix.
Watch out for this
Reliable data always give a valid conclusion.
A method can consistently measure the wrong thing.
Check your understanding
Every observer counts the same number of ramps, but ignores steep slopes. What is the main issue for judging wheelchair access?
- The indicator has limited validity for access.
- The count must be unreliable between observers.
- The sample of ramps is definitely too small.
Show the answer
The indicator has limited validity for access.
Yes. It is consistent but misses what matters.
Check your understanding
Students measure a river's width three times and get the same result each time, but they measured the wrong river for their question. Their data are:
- Both reliable and valid.
- Reliable but not valid.
- Valid but not reliable.
Show the answer
Reliable but not valid.
Repeatable results can still fail to measure what the question needs.