Evaluating the investigation
What the Excellence criterion asks
Excellence requires a comprehensive evaluation containing a selection from:
-
Evaluation of the reliability of the data, considering the procedure used and possible sources of error.
-
Justification of how the processed data supports the conclusion.
-
Linking the conclusion to chemical principles and/or real-life applications.
-
You do not need all three, but a strong report covers at least two — and the first is the one most likely to be done superficially.
Evaluating reliability
- Reliability is about whether the same procedure would give the same result again. Address it in three parts.
1. Precision — what the repeats show.
- Quote the spread of the concordant titres at each point, converted into a percentage of the result.
- Three titres within 0.10 mL of a 20 mL titre is about 0.5%, which is good precision.
- Then compare that with the size of the trend. If the trend is 50% and the scatter 0.5%, the trend is secure.
2. Systematic errors — what repeats cannot reveal.
- Name the specific ones that apply to your procedure, not a generic list:
| Error | Direction | How you addressed it |
|---|---|---|
| Titrant not standardised | either | standardised against a primary standard on the day |
| Conical flask rinsed with sample | high | rinsed with distilled water only |
| Analyte decomposing before titration | low | titrated promptly, kept cool and stoppered |
| Iodide air-oxidised in acid | high | stoppered, kept out of light, titrated immediately |
| Endpoint judged consistently late | high | same operator, same criterion, white tile |
| Colorimeter blank not matrix-matched | high | blank prepared from analyte-free matrix |
3. Validity — whether you measured what you intended.
- Did anything else in the sample react with your titrant?
- Ascorbic acid determinations by iodine titration are the classic case: other reducing substances in juice also reduce iodine, so the result is really "total reducing capacity" and will be high.
- Naming an interference like this, and saying which way it biases the result, is strong evaluation.
Justifying how the data supports the conclusion
- State the magnitude of the trend against the magnitude of the uncertainty.
- Say how many points support it and whether the pattern is consistent across them.
- Be explicit about what the data does not establish:
- A trend is not a mechanism. Falling ascorbic acid is consistent with oxidation but does not prove it.
- Correlation with time does not establish which time-dependent process is responsible.
Linking to chemical principles
- Explain the trend using the chemistry, with equations.
- Ascorbic acid declining on storage is oxidation by atmospheric oxygen — a redox process, faster in light and at higher temperature.
- Hypochlorite bleach weakening over time is decomposition, accelerated by light and warmth.
- Iron(II) supplements losing potency is oxidation of Fe2+ to Fe3+ by atmospheric oxygen.
- Then link to a real-life application: storage recommendations, labelling, shelf-life dating, best-before periods.
What a comprehensive evaluation is not
-
A list of everything that could go wrong, unranked and unquantified.
-
"Human error" — this means nothing. Name the specific action and its effect.
-
"The results could be improved by repeating the experiment" — true only for random error, and worth saying only if you also say which of your errors were random.
-
The distinguishing feature of a good evaluation is that it ranks the errors by how much they could have affected the result, and ties each to a specific step of your procedure.
Worked ExampleA comprehensive evaluation
Write an evaluation for an investigation that found the ascorbic acid content of orange juice fell by 57% over ten days of refrigerated storage, measured by iodine titration with three concordant titres per point at five time points.
Reliability of the data — precision
At every time point the three concordant titres agreed within 0.10 mL on titres of about 18–20 mL, a spread of roughly 0.5%. Converted through the calculation, this corresponds to an uncertainty of about ±0.7 mg per 100 mL in each result.
The observed change over the ten days was from 136 to 58 mg per 100 mL, a difference of 78 mg per 100 mL — approximately one hundred times the within-point spread. The trend is therefore far larger than the random scatter, and cannot be attributed to imprecision in the titrations.
Reliability of the data — systematic errors, ranked
Most significant: interference from other reducing substances. Iodine oxidises any reducing agent present, not only ascorbic acid. Orange juice contains other reducing species — sugars in small part, and various polyphenols — which also consume iodine. This makes every result too high, and it explains why the day-0 value of 136 mg per 100 mL is well above the 30–60 mg per 100 mL typical of fresh orange juice.
Crucially, this error affects the absolute values but probably not the trend, since the interfering substances are present at all time points. So the conclusion about the direction and approximate size of the change survives, while the absolute concentrations should be regarded as upper limits. Separating what the error invalidates from what it does not is the key judgement here.
Second: titrant drift. Iodine solution is not a primary standard — iodine sublimes and iodide is air-oxidised. It was standardised against potassium iodate on each day of use, which converts this from an uncontrolled error into a measured quantity. Had it not been, drift over ten days would have mapped directly onto the independent variable and produced an apparent trend indistinguishable from the real one.
Third: time between opening the sample and titrating. Ascorbic acid is oxidised by atmospheric oxygen, so any delay lowers the result. The delay was held to under ten minutes for every sample, so the error is small and, being constant across samples, does not distort the trend.
Fourth: endpoint judgement. Starch was added only when the solution had faded to pale straw, so the endpoint was sharp. The same operator judged every endpoint against the same criterion, making any bias constant across samples.
Justification of how the data supports the conclusion
Five time points all follow the same pattern, with the decline steepest in the first three days. The change is two orders of magnitude larger than the within-point scatter, and the run order was randomised so that titrant drift could not correlate with storage time.
What the data does not establish: the trend shows that ascorbic acid concentration falls, but does not identify why. Any process consuming ascorbic acid — oxidation by dissolved oxygen, microbial action, or reaction with other juice components — would produce the same shape. Attributing it to atmospheric oxidation is an inference from known chemistry, not a finding of this investigation. Testing it would require comparing juice stored under nitrogen against juice stored in air.
Linking to chemical principles and applications
Ascorbic acid is a reductant: it is oxidised to dehydroascorbic acid, losing two electrons and two protons. This is why iodine can be used to measure it:
The same property makes it vulnerable during storage: dissolved atmospheric oxygen oxidises it by the same route, which is why the concentration falls once the container is opened and headspace air is introduced. The decline being steepest early is consistent with the initially higher concentration driving a faster rate, and with the dissolved oxygen introduced at opening being consumed over the first few days.
Applications: this supports the standard advice to consume juice within a few days of opening, to keep it refrigerated (lower temperature, lower rate), and to store it in a full, sealed container to minimise headspace oxygen. It also explains why manufacturers add ascorbic acid to some juices in excess of the natural content — the added margin compensates for losses during shelf life, and may explain the high day-0 value measured here.
Overall judgement: the trend is securely established and its size is reliable to within a few percent, but the absolute concentrations are systematically high because iodine titration measures total reducing capacity rather than ascorbic acid specifically. The mechanism is consistent with oxidation but is not demonstrated by this investigation.