IB Chemistry SLTopic 8 — Collecting and Processing DataInternal assessmentPractical skill~13 min read
Interpreting Results
A graph is not decoration you add once the writing is finished. It is the argument. This page is about making the graph say something, and then explaining why the chemistry made it say that — because the second half is where nearly all the marks live.
📚 What you need to know
A graph needs a specific title, the independent variable on the x-axis, and axes labelled with quantity and unit.
The scale must be linear and sensible, filling at least half the space you have.
A line or curve of best fit shows the trend. It is never dot-to-dot.
Gradient, intercept and area can each carry chemical meaning — look before you ignore them.
Error bars show precision; a best-fit line passing through them all supports the fit.
Interpreting is two steps: describe the trend, then explain it with chemistry.
Accuracy, precision, reliability and validity are four different words. Do not swap them.
What a graph has to have
These requirements are dull and they are also free marks, which is a rare combination. Work through them once and the graph is done.
The scatter is deliberate. Real points do not sit perfectly on a line, and a best-fit line that visits every point exactly is usually a sign the data was tidied.
A common instinct is to force the line through the origin because it looks neater. Do not. Draw the best fit the points actually support, and then comment on where it crosses. If it should pass through the origin and does not, that gap is evidence of a systematic error — which is a far more interesting thing to write about than a tidy line.
Reading more than the shape
Most students describe the line going up and stop. The graph has more in it than that, and each of these features is a place to earn something.
Feature
What it can tell you
The gradient
Often a quantity in its own right. On a rate against concentration graph it is a rate constant; on a graph of moles against volume it is a concentration.
The y-intercept
The value of the dependent variable when the independent variable is zero. If that ought to be zero and is not, suspect a systematic error.
The area under the curve
A total rather than a rate — for example the number of particles under part of a Maxwell–Boltzmann distribution.
Error bars
The precision of each point. If the best-fit line passes within every bar, the trend is well supported by your data.
Where the curve flattens
Something has run out or become limiting. A plateau is a chemical event, not the graph getting bored.
Describe, then explain
Interpretation is two separate jobs and they get written as two separate sentences. Mixing them together is how a good result ends up sounding vague.
Useful words for the describing sentence: directly proportional, linear positive correlation, inversely proportional, exponential, and plateaus. Choose the one that is actually true.
Anomalies, and how to justify one
An anomalous result is a point that clearly does not belong to the trend. Circle it on the graph, leave it out of the best-fit line, and then do the part that most students skip: say what caused it.
“This point was anomalous” is not a justification, it is a label. A justification names a mechanism: the stopwatch was started late, the water bath had not yet reached temperature, a drop of titrant was left hanging on the burette tip. Link the point to a specific, plausible error and you have made an argument.
Four words that are not synonyms
Accuracy, precision, reliability and validity get used interchangeably in ordinary speech and it costs marks in chemistry, because each one is asking a different question about your work.
Repeating an experiment tightens the cluster. It does not move it. That is why doing more trials cannot cure a systematic error — only changing the method can.
Word
The question it answers
Affected by
Accuracy
How close is my result to the accepted value? You can only comment if a literature value exists.
systematic errors
Precision
How close are my repeats to each other?
random errors
Reliability
Would the same thing happen again? Concordant repeats make results reliable.
random errors, repetition
Validity
Was the method a fair test of the question I asked?
uncontrolled variables
Percentage error against a literature value
% error = |experimental − literature| ÷ literature × 100
Compare that percentage error with your total percentage uncertainty from the last page. If the error is smaller than your uncertainty, your result agrees with the literature within experimental limits — say so. If the error is much bigger, something systematic is going on and no amount of extra repeats will remove it. That single comparison is one of the strongest sentences you can write in an IA.
WORKED EXAMPLE
A graph of initial rate against HCl concentration gives a straight line passing through the origin. Describe and explain the trend.
DescribeThe initial rate is directly proportional to the concentration of hydrochloric acid: the line is straight, positive and passes through the origin, so doubling the concentration doubles the rate.ExplainA higher concentration means more acid particles in the same volume, so collisions between particles happen more often. Because the temperature is unchanged, the proportion of collisions with at least the activation energy stays the same — so twice as many collisions means twice as many successful ones.describe the line, then name collision theoryThe extra sentence worth addingThe line passing through the origin says that with no acid there is no reaction, which is exactly what you would expect. That agreement is evidence there is no large systematic error in the timing.
WORKED EXAMPLE
Three trials give an enthalpy of neutralisation of −52.1, −52.4 and −52.3 kJ mol−1. The literature value is −57.3 kJ mol−1. Comment on the precision, reliability and accuracy of the result.
PrecisionThe three values span only 0.3 kJ mol−1, a very small spread, so random error was low.preciseReliabilityThree concordant repeats agreeing this closely means the result is repeatable.reliableAccuracymean = −52.3; % error = (57.3 − 52.3) ÷ 57.3 × 100 = 8.7%Every value is less exothermic than the literature figure, and by a similar amount each time.precise but not accurateWhat that combination meansAn error in the same direction every time is systematic, and the obvious candidate is heat lost to the cup, the thermometer and the air. More repeats will not help. Insulating the calorimeter, adding a lid, or extrapolating a cooling curve back to the moment of mixing will.
WORKED EXAMPLE
On a graph of rate against temperature, the point at 40 °C sits well above the smooth curve formed by the other four points. Write the sentences an assessor is looking for.
Identify it on the graphCircle the point and say in the text that it was excluded when drawing the curve of best fit.Name a mechanism, not a moodFor example: the water bath had only just reached 40 °C, so the mixture was probably warmer than recorded when the reaction started, giving a rate that was too high.Classify the errorThis is a one-off random error rather than a systematic one, because the remaining four points sit on a smooth curve.Say what you would changeleave the flask in the bath until the reading holds steadyNotice the shape of that answer: identified, explained, classified, fixed. A sentence that only says “this was an anomaly” does the first quarter of the job.
💡 Exam tip
Give the graph a title naming both variables, not just “Results”.
Independent variable on x, dependent on y, both with quantity and unit.
Write the description and the explanation as two sentences, in that order.
Make the explanation name a principle: collision theory, bond enthalpies, intermolecular forces.
Justify every anomaly with a specific likely error, then say what you would change.
Compare your percentage error with your percentage uncertainty and comment on what that tells you.
⚠️ Common mix-up
Using precise to mean accurate. A tight cluster can sit a long way from the truth.
Describing the trend and stopping, leaving the chemistry unsaid.
Joining the points dot-to-dot instead of drawing one smooth best fit.
Claiming accuracy with no literature value to compare against.
Calling a result anomalous without naming a cause.
Suggesting more repeats as the cure for a systematic error.
Up next: Concluding and Evaluating — you have a result, an uncertainty and a trend you can explain. The last stage is judging honestly how much your method can be trusted, and saying what you would do differently.
Want this explained one-to-one?
Book a free session with an experienced IB Chemistry tutor and get your trickiest topics made simple.
