Repeatability means that repeating the same method gives similar results. Validity means that the method really tests the question and that the conclusion is justified. A result can have one without the other.
This lesson builds on choosing a control, because controlling variables is what makes an investigation valid.
How are the two ideas different?
Repeatability is a question about the data: do the repeated readings agree? You answer it by looking at the spread of repeats or the range.
Validity is a question about the design: was only the intended factor changed, and was the right thing measured? You answer it by checking control variables and the measurement itself.
How to check each one, step by step
- For repeatability: list the repeat readings for one condition and find the range. A small range means good repeatability.
- Calculate the mean of the repeats if they agree, and note any anomaly.
- For validity: write the independent variable, then read the method for anything else that changed between set-ups.
- Check the measurement: does it really measure what the question claims, such as oxygen produced, not just a bubble count of uneven size?
- Write both verdicts separately with one reason each.
Worked example
All data are invented for practice. A student investigates the effect of temperature on yeast by counting bubbles of carbon dioxide per minute. At 30 °C she repeats the count three times.
| Repeat | Bubbles per minute |
|---|---|
| 1 | 31 |
| 2 | 33 |
| 3 | 32 |
Step 1, repeatability: the readings range from 31 to 33, a range of 2 bubbles per minute.
Step 2, mean: (31 + 33 + 32) ÷ 3 = 96 ÷ 3 = 32 bubbles per minute. The readings agree closely, so the results are repeatable.
Step 3, validity check: reading the method, the tube at 30 °C contained 5 g of yeast but the tube at 50 °C contained 8 g. The mass of yeast changed along with the temperature.
Step 4, verdict: the results are repeatable but not valid for the question about temperature, because the mass of yeast was not controlled. A higher count could come from more yeast rather than higher temperature.
Step 5, the fix: use the same mass of yeast in every tube, then measure again.
Contrast a second student whose three repeats were 20, 35 and 26 bubbles per minute. The range is 15 and the mean is (20 + 35 + 26) ÷ 3 = 81 ÷ 3 = 27. The results are poorly repeatable, and the first step is to find out why the counts vary.
The mistake to watch for
The usual mix-up is to call a fair test repeatable, or to call consistent results valid.
Mistaken answer: “The results are valid because the three readings were almost the same.”
Agreeing readings show repeatability. Validity depends on whether the variables were controlled.
The correction is to use each word for its own job: repeats agree means repeatable, other variables controlled means valid. If you are asked for one, do not substitute the other.
Check yourself
All data are invented.
1. Four plants in a light experiment gave leaf counts of 6, 6, 7 and 6 under the same lamp. What can you say about repeatability?
Show answer
The readings have a small range of 1 leaf, so the results are repeatable. Repeatability does not say anything about whether other variables were controlled.
2. A student compares growth in sunlight and in shade, but the shade plants were given more water. Are the results valid? Explain.
Show answer
Not valid. Water was not kept the same, so growth differences could be due to water as well as light. Use the same volume of water for every plant.
3. An investigation has a range of 15 between repeats. Suggest one reason it is poorly repeatable and one check.
Show answer
The method may be hard to carry out consistently, for example counting bubbles of uneven size by eye. Check by repeating with a more precise measurement, such as collecting gas in a syringe, and see whether the range falls.
Where this leads next
The last lesson in this module is turning a vague improvement into a specific measurement change. The mixed practice then tests all five skills together.
If you can explain the two words aloud but still swap them on paper, online one-to-one Biology tuition can train the exact wording with your own answers. The module overview lists every lesson.