IB Physics SLInquiry 2 — Collecting & Processing DataInternal AssessmentRaw data tables & observations~8 min read
Collecting Data
You’ve designed a solid experiment — now comes the doing. This is the hands-on phase where you run the method and gather your raw data. Get it right and everything downstream (your graphs, your conclusion, your marks) rests on firm ground. Get sloppy here and no amount of clever analysis can rescue it. The goal: data that’s both accurate and sufficient to actually answer your research question.
📘 What you need to know
Your report’s foundation is a raw data table — only direct measurements, no calculations
Design the table before you start, with units and uncertainties in the headers, not the body
Record every reading to the precision of the instrument (e.g. 0.550 m, not 0.55 m)
Sufficient data means five or more values of the IV, each with three or more repeat trials
Qualitative observations are evidence too — record what you see, hear, and feel
Notice and respond to problems as they happen; never quietly ignore or erase them
The raw data table
Everything begins with one table. The raw data table is the very first thing you present in your report, and it contains only the numbers you read straight off your instruments — no processed values, no averages, no calculated results. Those come later, in their own table.
The single most important habit: design the table before you touch the apparatus. A well-built table forces you to think about exactly what you’ll measure, in what order, and how many times — before a rushed lab session makes you improvise.
🧭 What every raw data table must have
A specific title — describe the actual experiment, not just “results”.
Clearly labelled columns — independent variable first, then the dependent variable(s).
Units and uncertainties in the headers — write them once, at the top, using the ± symbol.
Consistent precision — every value in a column has the same number of decimal places.
Here’s the rule students trip over most: units and uncertainties live in the column header, never in the body of the table. You write “Length / m (±0.001)” once at the top — then the cells below just hold bare numbers like 0.200, 0.400, and so on. It keeps the table clean and is the proper scientific convention.
Units and uncertainties belong once in the header — the table body holds bare numbers with consistent decimal places.
Recording to the right precision
Your readings must match the precision of the instrument — this is a classic place to drop easy marks. If a metre ruler is marked in millimetres, a reading of 55 cm is written as 0.550 m in metres (three decimal places), not 0.55 m. If a digital multimeter shows two decimal places, then a five-volt reading is 5.00 V, not 5 V. Those trailing zeros aren’t decoration — they tell the reader how precisely you measured.
Quick recap: record what the instrument can actually resolve. A digital scale reading to 0.01 g gives you 12.30 g, never 12.3 g.
What “sufficient” data means
Enough data to see a genuine trend. As a working rule for an IA, aim for at least five different values of your independent variable, and at least three repeat trials at each one. The five values let a pattern emerge; the repeats let you spot anomalies and average out random error.
5+ IV values
→ reveals →
a clear trend
3+ repeats
→ enables →
reliable average
Qualitative observations count too
Qualitative data is the non-numerical stuff you notice while the experiment runs — and it’s genuinely useful, not filler. It provides the context that later explains a surprising result. Don’t treat it as an afterthought.
Physics examples worth recording: a pendulum bob that swings in a slight ellipse instead of a flat plane; a digital meter reading that flickers rather than settling; a wire or resistor that feels warm to the touch, hinting at heating; a laser beam visibly spreading as it passes through a narrow slit. Any of these can later explain why your numbers drifted from the theoretical value.
Here’s the mindset shift that separates a good IA from an average one: your observations are evidence. When your calculated resistivity comes out too high and you can point to “the wire felt warm after several readings”, you’ve handed yourself a ready-made explanation in the evaluation. Examiners reward that link between what you saw and what you calculated far more than a table of numbers alone.
Handling problems as they happen
Experiments rarely run perfectly, and noticing issues in real time is a real scientific skill. The golden rule: if something goes wrong, don’t ignore it — record it in your lab notes and record how you responded.
🧭 Three common issues and how to respond
Unstable reading — if a voltmeter won’t settle, wait for it to stabilise, or record the central value and note the fluctuation as an uncertainty.
An anomalous result — if one repeat is wildly off, don’t erase it. Record it, then run an extra trial to get a reliable set. You’ll justify excluding it later.
Difficulty measuring — if a length is hard to read because it’s oscillating, note it as a limitation to discuss in your evaluation.
WE 1
A student times 20 oscillations of a mass on a spring, three times, for one spring length. Their stopwatch reads to ±0.2 s.
Trial times for 20 oscillations: 22.6 s, 22.8 s, 22.5 s.
(a) Find the average time for 20 oscillations. (b) Later they will divide by 20 to get the period — state the percentage uncertainty in that time.
Part (a) — average of the raw times
add the three trials and divide by 3
t₀ = (22.6 + 22.8 + 22.5) ÷ 3 = 22.63 st₀ = 22.6 sThis average is a processed value — the three raw times stay in the raw table.Part (b) — percentage uncertainty
% uncertainty = (absolute ÷ measured) × 100
= (0.2 ÷ 22.63) × 100 = 0.88%≈ 0.88%Dividing by 20 (an exact number) doesn’t change this — the period carries the same 0.88%.
WE 2
For one length of resistance wire, a student records the raw readings below. State the qualitative observation they should log, and how they’d handle a fluctuating meter.
Potential difference V = 1.68 ± 0.01 V, current I = 1.24 ± 0.01 A. The wire felt warm after several readings.
Raw data — what goes in the table
record V and I directly, to the meter’s precision (2 d.p.)
The resistance R = V ÷ I is processed data — it does NOT belong in the raw table.Qualitative observation to log
“the wire felt warm to the touch after several readings”
This is evidence: heating raises resistance, which can explain a high calculated value later.Handling a fluctuating meter
wait for it to settle, or record the central value
and note the spread as uncertainty
💡 Top tips
Record raw data directly — never do a calculation in your head or on scrap paper. The table shows what you measured, not what you worked out.
Units and uncertainties belong in the headers — write them once at the top, keep the cells bare.
Your observations are evidence — a warm wire or a wobbling stand can explain your results in the conclusion, so log them.
Keep decimal places consistent down every column — 0.200, 0.400, 0.600, matching the instrument’s precision.
⚠ Common mistakes
Putting units next to every number in the table body instead of once in the header
Dropping trailing zeros — writing 5 V or 0.55 m when the instrument justifies 5.00 V or 0.550 m
Sneaking calculated values (like resistance or period) into the raw data table
Erasing an anomaly instead of recording it and running an extra trial
Up next: Processing Data — turning those raw numbers into averages and final values, propagating uncertainties through your calculations, and presenting it all in clear tables and graphs.
Want this to actually click before the exam?
Book a free meeting and let’s work through the tricky bits together.
