The Fifth Reading
The practical examination in an Intermediate science subject carries 30 marks, and 24 candidates are scheduled between 10:00 a.m. and 4:00 p.m. Ananya is performing an experiment that requires five concordant burette readings. Her rough sheet shows 24.1, 24.2, 24.1, 24.3 and then 26.8 mL. She strikes out the 26.8, writes 24.2 in its place, and adds a footnote: air bubble at the tip, reading rejected and re-taken. Her fair record contains nine experiments, each with aim, apparatus, observations and result, but in Experiment 7 the column headings carry no units and the final calculation is out by a factor of ten; the correction is entered in red ink and initialled. The external examiner reaches Ananya at 2:40 p.m. Eight candidates are still waiting, and the laboratory assistant has already begun packing the apparatus. The marking grid allots 10 marks to the experiment, 5 to the viva, 5 to the record, and the remaining 10 to the other components of the session.
1. Should the struck-out 26.8 mL reading count against Ananya, be ignored, or be treated as a serious problem? Set out the reasoning that separates an honest re-take from a fabricated one, and state what further evidence on the rough sheet would move your judgement either way.
Read the worked answer
A practical examination is testing whether the candidate can recognise and handle a faulty reading, so rejecting one is the behaviour being rewarded, not penalised. What separates honest from fabricated is whether the correction leaves a trail and whether the replacement is defensible.
Here the trail is good: the original 26.8 is still legible, a physical cause is named, and the replacement 24.2 sits inside the cluster of 24.1 to 24.3 rather than being suspiciously exact. An air bubble would displace liquid and then, once removed, give a value consistent with the others, so the note and the number agree.
I would ignore the episode: no penalty, but no extra credit either.
Evidence that would change this: if the 26.8 had been overwritten so the original could not be read, or if the replacement made the mean land exactly on the theoretical value, I would treat it as reverse-engineered. If the rough sheet showed three or four such corrections across the same experiment, I would ask for the readings to be re-taken under observation.
2. The record carries 5 marks. Award a mark for Ananya's record and justify the deduction, stating also what you would refuse to accept even if the rest of the record were flawless.
Read the worked answer
The record mark is for sustained, honest record-keeping, not for a perfect notebook. Nine experiments are present with aim, apparatus, observations and result, so the completeness requirement is met and the starting position is not zero.
Experiment 7 has two defects. Missing units in a column heading is carelessness that propagates into the result, and the factor-of-ten error is a calculation failure. But the error was caught and the correction is in red ink and initialled, which is exactly how a working record should show a correction. I would deduct 1 mark and award 4 out of 5, the deduction resting on the unit-less table because a heading error affects every entry beneath it.
What I would not accept: a correction with no initial and no date, because then the examiner cannot tell whether the candidate or a later hand wrote it; and a result that contradicts the observation table it is supposed to follow. If the unit omission appeared in three or more experiments, I would deduct 2 rather than 1, since that is a systematic gap in the record and not an isolated slip.
3. Write the entry you would place against Ananya's record mark on the award sheet, and explain why that wording matters when 23 other candidates are being marked the same afternoon.
Read the worked answer
Entry: Record 4/5. Nine experiments complete. Experiment 7 column headings lack units and the result carries a factor-of-ten error, corrected in red and initialled. Deduction applies to the unit omission only.
The wording matters for three reasons. First, a bare number is unreviewable: if the mark is questioned, the note records the observable feature and the rule applied, not an impression. Second, it protects the examiner against the charge of arbitrariness, because the reason can be checked against the script by anyone.
Third, and most practically, writing the reason forces consistency across the remaining candidates. The rule I stated for Ananya - deduct 1 for a unit-less table, 2 if it recurs in three or more experiments - has to be applied to the eight candidates still waiting and to the 15 already marked. If I discover at 3:30 p.m. that I have been deducting 2 for a single unit-less table, I must go back rather than let the standard drift with the clock.