Graphs and data processing
In a minute
A graph turns a table of results into a pattern you can read. You'll spot and , draw a , judge the fit with , and read rates from it. show how far each point could move.
You'll be able to
- Recognise direct and inverse proportionality, and positive and negative correlation.
- Sketch graphs with labelled axes, and plot graphs with sensible scales and a line or curve of best fit.
- Use to judge how well a trend line or curve fits.
- Find rates from tables, and interpret the , intercepts, maxima, minima and areas of graphs.
- Draw and interpret uncertainty bars, and use and .
Before you start3 quick questions on what this builds on
Revise first: 5.3 Rates of reaction and collision theory and 7.2 Uncertainties, errors and significant figures
Variables, proportionality and correlation
The is the one you change on purpose. The is the one you measure. Everything else that could affect the result is kept the same.
To test a table, check whether or is constant. A straight line that misses the origin, such as , is linear but not .
A correlation is a looser link. In a positive correlation tends to rise as rises; in a negative one it tends to fall. A correlation alone doesn't prove that one variable causes the other to change.
Try it first, then reveal one step at a time.
- falls as rises. is then , so it is not direct proportionality.
- : , , , . Constant, so is inversely proportional to .
- At : .
Try it first, then reveal one step at a time.
- Ratios: , , . Not constant, and the ratios in kelvin aren't constant either.
- At the solubility is , not zero.
- So the trend is a correlation only.
- There is a positive correlation: solubility increases as temperature increases. It is not direct proportionality, because the ratio of solubility to temperature is not constant. The solubility at 0 °C is 13 g, not zero, so the graph does not pass through the origin.
Sketching, plotting and lines of best fit
A sketch shows the shape of a trend: both axes labelled, no scale. Show what matters, such as where the line starts, whether it curves, and any maximum.
A plot is drawn to scale. Put the on the -axis. Label each axis "quantity / unit", such as "time / s".
Choose scales so the points cover at least half the grid each way, using easy steps (1, 2 or 5 units per square, times a power of ten). An axis need not start at zero.
Don't join the dots: the scatter comes from . If the points curve, draw one smooth curve. Go through the origin only if is a real result.
A point far from the trend is anomalous. Circle it and leave it out of the .
Try it first, then reveal one step at a time.
- Temperature is changed, so it goes on the horizontal axis.
- "Sketch" means labelled axes, no scale.
- Volume is directly proportional to kelvin temperature.
- Volume is on the vertical axis and temperature / K on the horizontal axis, labelled with no scale. The graph is a straight line rising from the origin, because volume is directly proportional to temperature in kelvin.
Try it first, then reveal one step at a time.
- Ratio of oxide to magnesium: , , , , . The result is anomalous.
- The good points lie on a line through the origin with gradient .
- Expected mass , so about .
- A likely cause: oxide escaped as smoke when the lid was lifted.
How well a trend line fits: $R^2$
Spreadsheets can add a trend line or curve to data. They also give , a number that measures how closely the points follow it.
Use to compare trend lines for the same data: the one with nearer 1 fits better. For a trend of the right shape, a low value means scattered points, usually from random errors, so poorer .
A high doesn't prove cause, and it doesn't show . A moves every point the same way and can leave unchanged.
Try it first, then reveal one step at a time.
- Compare each with 1.
- Points above the line at both ends and below it in the middle mean the data bend.
- The exponential curve fits better. Its of 0.996 is closer to 1 than 0.914, so the points lie closer to it. The pattern of points about the straight line shows the data are curved.
Try it first, then reveal one step at a time.
- Absorbance should be proportional to concentration, so a straight line is the right shape. Both values are close to 1, so it fits both sets well.
- B's points scatter more.
- Neither value tells you about accuracy.
- A straight line fits both data sets well, but group A's points lie closer to its line. Group B's points are more scattered, which suggests larger random errors and lower precision. cannot show accuracy, because a systematic error could shift every point without changing .
Rates, gradients and other features
A rate of change is how fast one quantity changes, usually per unit time. From a table, subtract two rows. This gives the average rate over that interval, not the rate at one instant.
Find a from two points far apart on the line itself. On a curve, use the at that point (5.3). On a graph against time, a curve getting less steep means the rate is falling; a flat section means it is zero.
The -intercept is when , such as a starting temperature. A maximum or minimum marks where a trend turns round, such as the peak temperature in a .
The area under a graph is the quantity times the quantity, added up. Under a curve it is the total number of particles.
Try it first, then reveal one step at a time.
- Change in volume: . Time: .
- Average rate , so about .
- The next give only : the rate falls as the peroxide is used up.
Try it first, then reveal one step at a time.
- Neutralization is exothermic.
- The peak is where the acid runs out.
- After that, nothing reacts.
- Neutralization is exothermic, so the temperature rises while the alkali reacts with the acid. The maximum is reached when all the acid has reacted, so 18.0 cm³ is the equivalence point. After this no more heat is released, and the added cooler solution and heat loss make the temperature fall.
Uncertainty bars, interpolation and extrapolation
A value of gets an from to . Convert a to an one first.
is usually reliable. is less so: a calibration line, for example, often curves at high concentration.
In , extend the cooling line back to the moment of mixing (for an endothermic reaction, the warming line). That estimates the maximum (or minimum) temperature with no heat exchange with the surroundings.
Try it first, then reveal one step at a time.
- Absolute uncertainty: .
- The bar runs from to .
Try it first, then reveal one step at a time.
- Extrapolate to : .
- . A change of is a change of .
Common mistakes
- Joining the dots. Draw one line or smooth curve with the points evenly on both sides.
- Forcing a line through the origin when is not a real result.
- Calling a rising straight line directly proportional when it misses the origin.
- Giving a gradient with no unit, or using data points that sit off the line.
- Reading a high as proof of accurate results or of cause and effect.
Exam tips
- "Sketch": right shape, labelled axes, no scale. "Plot": scaled axes with units, accurate points and a line of best fit.
- "Determine" a rate: show the two points used and the subtraction, then the value with its unit.
- In Paper 1B data questions, say whether the line passes through the uncertainty bars, and name any anomaly.
Key takeaways
Practise this
42 past-paper questions on Graphical and data processing in the Chemistry HL bank, each with a mark scheme and a worked explanation.