Revise Reliability, Validity & Suggesting Improvements for Biology 3 (IAL) WBI13 (AS Level) — revision notes and instant AI marking. Free to start.
📖 Revision notes · preview
Edexcel IAL Biology • Unit 3: Practical Skills in Biology I
Reliability, Validity & Suggesting Improvements
🔍 Big idea: Reliability and validity are not the same thing, and the improvement that fixes one will not fix the other. A reliable experiment gives the same answer every time; a valid one measures what it claims to. You can be perfectly reliable and completely invalid.
Summary — What This Topic Covers
What reliable means, and how repeats improve it
What valid means, and why controlled variables decide it
Why an experiment can be reliable but not valid
Writing improvements that are specific enough to score
The difference between an improvement and an extension
1. Reliability
Key Term
A reliable result is one that would be obtained again if the experiment were repeated under the same conditions. It is about the consistency of the measurement, not its truth.
Judged by how closely the repeats agree with each other
Improved by taking more repeats and calculating a mean
Improved by reducing random error — finer apparatus, more consistent technique
Shown on a graph by short error bars and little scatter about the line
How to describe it
"The three repeats at 30 °C ranged from 13.5 to 14.5 cm³, a spread of only 1 cm³, suggesting the results are reliable." Quote the spread — that is the evidence.
2. Validity
Key Term
A valid experiment measures what it set out to measure. It requires that only the independent variable changed, and that the dependent variable genuinely reflects the thing being investigated.
Validity is decided by the controlled variables. If pH drifted while you varied temperature, any change in rate could have come from either, and the experiment cannot answer the question it asked — however consistent the numbers were.
Every relevant variable other than the IV must be held constant
A control is needed to show the effect is caused by the IV and not by the procedure itself
The measurement must actually reflect the biology — counting bubbles is a poor proxy for oxygen volume, because bubble size varies
The one they test
An experiment can be reliable but not valid: three repeats agreeing perfectly, all measuring the wrong thing. It can never be valid but unreliable in any useful sense — if you cannot reproduce it, you cannot claim it measured anything.
3. Controls
Key Term
A control is a run in which the independent variable is absent or inactive, everything else being identical. It shows what would have happened anyway.
INVESTIGATION CONTROL
─────────────────────────────────────────────────────────────────
effect of enzyme on starch boiled (denatured) enzyme
effect of light on rate tube wrapped in foil
effect of a solution on mass distilled water, 0 mol dm⁻³
effect of an antibiotic disc disc soaked in sterile water
Note that a good control is not simply "leave it out" — it is the same procedure with the active factor removed. Boiled enzyme controls for the volume, the protein and the liquid; omitting the enzyme entirely does not.
4. Improvements That Score
The structure
Name the problem, then the specific change, then what it fixes. Three parts, as with risk assessment.
✗ "Be more accurate."
✗ "Do more repeats." ← only right if the problem was random error
✗ "Use better equipment."
✓ "The endpoint was judged by eye, which is subjective and gives random
error. Using a colorimeter to measure absorbance would give an
objective, repeatable endpoint."
✓ "Temperature was not controlled, so it may have affected the rate
alongside pH, making the result invalid. Using a thermostatically
controlled water bath at 30 °C would remove this."
Match the fix to the fault
More repeats fixes reliability. Calibration fixes systematic error. Controlling a variable fixes validity. Offering the wrong one shows you have not diagnosed the problem.
5. Improvement vs Extension
The difference
An improvement makes the same investigation better. An extension asks a new question. Read which one is wanted — answering the wrong one scores nothing.
QUESTION: effect of temperature on the rate of an enzyme
IMPROVEMENT use a water bath rather than a beaker of warm water, so
the temperature stays constant throughout each run
EXTENSION repeat at a different pH, to find whether the optimum
temperature changes with pH
Practice Questions
Practice Question 1
Explain the difference between reliability and validity, and give an example of an experiment that is reliable but not valid.
Practice Question 2
A student investigating the effect of an enzyme on starch uses a tube with no enzyme as a control. Suggest a better control and explain why.
Practice Question 3
A student concludes that their results are unreliable because their equipment had a zero error. Comment on this statement.
Practice Question 4
Suggest one improvement and one extension for an investigation into the effect of light intensity on the rate of photosynthesis in pondweed.
What to Memorise
Reliable = repeatableValid = measures what it claimsReliable ≠ validRepeats → reliabilityControlled variables → validityCalibration → systematic errorControl = same procedure, factor removedImprovement = same question, betterExtension = new question
Concepts Checklist
Exam Tips
What mark schemes look for
Improvements are marked on specificity. Name the apparatus or the variable, and say what it fixes. Comparatives on their own — "more accurate", "better" — score nothing.
The trap
Offering "more repeats" for a systematic error. It is the right answer to a different question, and examiners use it to separate students who have understood from those who have memorised.
Easy marks
Quote the spread of your repeats as evidence for reliability. "The three values ranged over only 1 cm³" is worth more than the word "reliable" on its own.
Worth remembering
Read whether the question wants an improvement or an extension. They are different questions and the wrong one scores zero however good it is.