9.7 - Evaluating a Method
Evaluation questions
- Evaluation questions ask you to judge how well an experimental method worked by identifying a weakness and suggesting an improvement.
- These often appear as paired questions: you must describe what caused an inaccurate result and how to fix it.
Example question: In an experiment burning a crisp under a test tube of water, the energy absorbed by the water is lower than expected. Suggest why.
Answering paired evaluation questions
- Source of error - Name a specific weakness that genuinely applies to the method used, rather than a generic issue like human error.
- Improvement - Suggest a realistic modification to the apparatus or method that specifically reduces that error.
A typical question
In the experiment, the amount of energy absorbed by the water is lower than the expected value. Suggest why the value in the experiment is lower than expected.
Sources of error and improvements
A source of error is a flaw in the experimental setup or method that causes measured values to be higher or lower than the true value.
Common errors and improvements in chemistry practicals
| Source of error | Why it affects the result | Improvement that fixes it |
|---|---|---|
| Heat lost to the surroundings during calorimetry (e.g. burning a food sample under water) | Not all released energy reaches the water, so the measured temperature rise is too low | Add insulation, such as a lid or a draught shield around the flame |
| Gas escaping before sealing the container when measuring reaction rates | Gas produced before inserting the bung escapes, so the measured volume is too low | Use a divided flask or suspend one reactant in a small tube inside the flask, mixing only after sealing |
| Judging a titration end point by eye | Over- or under-shooting the exact colour change makes the measured titre inaccurate | Perform a rough titre first to locate the approximate end point, then add solution dropwise near that volume |
Concordance and repeating readings
- A single reading gives no indication of whether a result is reliable or anomalous.
- Titre results are concordant when they agree within 0.2 cm3 of each other.
- Only concordant results are used to calculate an average titre; rough titres and outliers are excluded.
Identifying concordant results
- Example 1 - A rough titre of 15.6 cm3, followed by careful titres of 14.9 cm3 and 15.1 cm3.
- The careful titres are within 0.2 cm3 of each other, so they're concordant.
- Example 2 - Titres of 16.4 cm3, 15.5 cm3, and 15.7 cm3.
- Only 15.5 and 15.7 cm3 are concordant and used for the average; 16.4 cm3 is discarded.
Representative sampling
- Where a substance naturally varies across different locations or batches (e.g. testing the pH of a river along its length or testing tablets from a production line), multiple samples must be taken.
- Calculating the mean of multiple samples ensures the result is representative of the whole, rather than reflecting just one local spot.
Common exam questions
Question 1
A student burns a crisp under a test tube of water and finds the energy absorbed by the water is lower than expected. Suggest why.
Heat is lost to the surroundings rather than all of it reaching the water, for example because the flame is not enclosed and the glass conducts heat away, so the temperature rise measured is smaller than the true value.
Question 2
Why does "heat is lost" on its own not fully answer a question that asks for an error and an improvement?
Because the two are judged together; you also need to say what you would change, such as insulating the container or shielding the flame, so less heat escapes before it reaches the water.
Question 3
A student adds acid to magnesium in a flask and only pushes the bung in once bubbling has started. Identify a source of error and an improvement.
Some hydrogen gas is produced and escapes before the flask is sealed, so the volume collected is lower than the amount actually produced; sealing the flask at the same moment the reactants are mixed, for example by tipping the last reactant in from a stoppered flask, would fix this.
Question 4
A student times a colour change in a titration by watching for it and stopping as soon as they think they see it. Give one source of error and an improvement.
Judging the end point by eye is a judgement call, so a drop too many or too few is easy to add before stopping; running a rough titration first to find roughly where the end point falls, then adding the solution drop by drop near that point on further titrations, reduces this.
Question 5
A student runs one titre and uses it to calculate a concentration. Why is this a problem?
A single titre gives no way of checking whether the result can be repeated; only once a second titre is run and compared can you tell whether the reading is concordant and safe to use.
Question 6
Titre volumes of 15.6, 14.9, and 15.1 cm3 are recorded. Which are used to calculate the average, and why?
The 14.9 and 15.1 cm3 titres, because they are within 0.2 cm3 of each other and so are concordant; the 15.6 cm3 titre is a rough result, run to find roughly where the end point falls, and is not close enough to the other two to be included.
Question 7
Why is "repeat the experiment" not a full answer to "identify a source of error and suggest an improvement"?
Because it does not name what actually went wrong on that run; a source of error has to be a specific weakness in the method, paired with an improvement that fixes that specific weakness, not a general instruction to do it again.