This is the doing part — the lesson where the design you spent a fortnight on finally meets a beaker. Two things decide whether it goes well: a results table you drew before you started, and the discipline to write down what actually happened rather than what you hoped would.
📘 What you need to know
Your raw data table holds only direct measurements — no calculations, no processed values.
Units and uncertainties go in the column headings, never beside every number.
Record to the precision of the instrument: 1.50 g from a two-decimal balance, not 1.5 g.
Sufficient data means at least five values of the independent variable and at least three repeats of each.
Qualitative observations are data too, and they often explain your odd numbers later.
When something goes wrong, record it and respond — do not quietly erase it.
Design the table before you start
Drawing your blank results table during planning is not admin. It is the last check on your design, and it catches things nothing else does — a missing initial mass, no column for the observation you will definitely want later, no room for a fourth repeat.
Every number in the body is something a person read off an instrument. That is the whole test of whether a table counts as raw data.
What makes the headings right
Quantity, unit and uncertainty, in that order: “time / s (±0.5)”.
Independent variable in the first column, dependent variable columns after it.
A specific title. “Results” is not a title; “time for starch to be broken down by amylase at five temperatures” is.
Recording to the right precision
This is a quiet, steady source of lost marks. The instrument decides how many decimal places you write, and every value in a column must be written to the same precision.
The trailing zero in 1.50 g is doing real work: it says the balance measured hundredths, and that the reading genuinely landed on zero.
Qualitative observations count as data
Your numbers say how much. Your observations often say why — and when a result surprises you months later, the observation is the only thing that can rescue it.
Texture and turgidity — plant tissue feeling firm and stiff, or soft and limp.
Colour — leaves yellowing, a solution clearing, a stain developing unevenly.
Behaviour — invertebrates moving away from a light source, or clustering in a damp corner.
Appearance of the reaction — froth forming, a suspension staying cloudy longer than expected, bubbles coming in bursts rather than steadily.
Write observations as you go, in the same place as your numbers. Nobody remembers on Thursday that the third tube looked cloudy on Monday, and “the solution never fully cleared at 50 °C” is exactly the sentence that explains an odd data point later.
When things go wrong — and they will
Biological experiments rarely run perfectly. Noticing a problem and responding to it sensibly is a scientific skill in itself, and it is credited. Ignoring it is not.
Marks for this come from the response, not from having a flawless run. A recorded problem with a sensible fix reads far better than suspiciously perfect data.
Three problems you should expect
The reaction is too fast or too slow. Adjust the enzyme or substrate concentration so the event is measurable, then record the change and why you made it. If this happens in a pilot, even better.
A repeat disagrees with its neighbours. Record it, run an extra trial at that value, and decide later whether to exclude it.
The organisms do not cooperate. Empty quadrats, seeds that never germinate, woodlice that will not settle. Change what you must, and write down the decision and your reasoning.
WORKED EXAMPLE
At 30 °C your three trials give 96 s, 148 s and 94 s. Describe what you should do during the lesson, and what you must not do.
Step 1: notice it
Trial 2 is 148 s against 96 s and 94 s — over 50 s adrift, far outside the spread of the other repeats.
Step 2: record it, with the likely cause
Note in your table that trial 2 may be anomalous, and add the observation: the enzyme was added before the tube had fully reached 30 °C.
Step 3: respond
Run a fourth trial at 30 °C. If it comes out near 95 s, you have three concordant readings and one outlier.
Record, repeat, and keep the original numberWhat you must not do is rub out 148 and pretend it never happened. You need it in the table to justify excluding it when you process the data.
WORKED EXAMPLE
Write two qualitative observations for the amylase experiment that would be genuinely useful later.
Observation 1 — at the low end
At 10 °C the mixture stayed a deep blue-black for a long time and cleared very gradually, so the end point was harder to judge than at higher temperatures.
Observation 2 — at the high end
At 50 °C the colour faded more slowly again, and the solution stayed slightly cloudy rather than going fully colourless.
Why these are useful
The first flags that readings at 10 °C carry more uncertainty. The second supports denaturation as the explanation for the falling rate.
Observations that will do work in your conclusion“It was hard to see” is a complaint. “The end point was harder to judge at 10 °C, so those readings are less certain” is evidence.
Take a photograph of your setup on the day. It costs nothing, and it settles arguments later about how far the lamp really was or which tube was which.
💡 Exam tip
Draw the blank table before the practical. It is the last chance to spot a missing measurement.
Units and uncertainties in headings; numbers only in the body.
Match every value in a column to the same number of decimal places.
Five values of the independent variable, at least three repeats of each.
Write qualitative observations alongside the numbers, not from memory afterwards.
Record problems and what you did about them — that is a credited skill, not an admission of failure.
Keep your raw table untouched. Everything you calculate goes somewhere else.
⚠ Common mix-up
Putting means or rates in the raw data table. Raw means measured, nothing else.
Writing units after every number in the table body.
Dropping a trailing zero, so 1.50 g becomes 1.5 g and precision is thrown away.
Copying the full calculator or sensor display when the instrument is not that precise.
Erasing an anomalous reading instead of recording it and running another trial.
Treating observations as optional, then having nothing to explain a strange result with.
Recording results on scrap paper and writing them up neatly later, losing detail on the way.
Up next: Processing Data — means, rates, percentage changes and standard deviations, and how to show your working so the calculations can be followed.
Want this explained one-to-one?
Book a free session with an experienced IB Biology tutor and get your trickiest topics made simple.
