IB Psychology HLTopic 5 — Data AnalysisPaper 3 & IAHL only~10 min read
Reading and Drawing Graphs
Choosing the wrong graph is the fastest way to lose marks in the HL paper, and it is completely avoidable. Each graph type belongs to a particular kind of data, and once you know which is which, the drawing part is easy. The gaps between bars are not a design choice — they carry meaning.
📚 What you need to know
Bar graphs show discrete or categorical data and always have gaps between the bars.
Histograms show continuous data and the bars touch, because the categories run into each other.
In a histogram, a category with zero frequency leaves a genuine gap — and that gap means something.
Line graphs show how a continuous quantity changes over time, with measurements taken at regular intervals.
Scatterplots show the relationship between two co-variables, one on each axis.
Every graph needs a title, both axes labelled with units, and a sensible scale.
The x-axis usually holds the categories or conditions; the y-axis usually holds frequency or score.
Bar graph or histogram? The gaps decide
This is the distinction examiners test most. A bar graph shows things that come in separate boxes — three experimental conditions, four eye colours, five shoe sizes. There is nothing between condition A and condition B, so a gap between the bars is honest. A histogram shows a continuous scale that has been chopped into intervals: ages, times, scores. Age 29 runs straight into age 30, so the bars have to touch.
Shoe size and eye colour get gaps. Age, time and test score do not. That single test settles almost every exam question on this topic.
One catch worth knowing: in a histogram, an interval with a frequency of zero leaves an actual empty space. That is not a mistake — it is data. It tells you nobody scored in that range.
Line graphs: change over time
A line graph shows how a continuous quantity moves across time. Measurements are taken at regular intervals, plotted as points, and joined with straight lines. The joining is what makes the pattern visible — a rise, a dip, a plateau — and the pattern is usually the whole point of the graph.
A second line can be added for a second data set — temperature and number of arrests on the same day, for example — but then the graph needs a key.
Which graph for which data
Graph
Use it when
x-axis
y-axis
Bar graph
Comparing separate categories or conditions
The categories or conditions
Score, mean or percentage per category
Histogram
Showing the distribution of one continuous variable
The measured intervals, in order
Frequency of each interval
Line graph
Showing change in a quantity over time
Time, at regular intervals
The measured quantity
Scatterplot
Showing a relationship between two co-variables
One co-variable
The other co-variable
The scatterplot rule you can rely on: it makes no difference which co-variable goes on which axis. A strong positive relationship looks the same either way round. That is not true of any other graph on this page.
🧩 Drawing a graph that scores full marks
Identify the data type first. Discrete or continuous? That decides bar versus histogram before you draw anything.
Title the graph with what it actually shows, not just “results”.
Label both axes including units. Unlabelled axes lose marks even if the shape is perfect.
Start the y-axis at zero unless you have a stated reason not to. Cutting the axis exaggerates differences.
Keep the scale even. Uneven intervals distort the shape and are treated as an error.
Add a key if there is more than one data set.
Worked examples
WORKED EXAMPLE
Choose the correct graph
A researcher records the mean recall score of three groups: one that revised with music, one in silence, and one with background chatter. Which graph should be used, and what goes on each axis?
Step 1: Classify the x-axis data
Three separate conditions — there is nothing in between “music” and “silence”, so the data are categorical.
Step 2: Choose the graph
A bar graph, with gaps between the bars.
Step 3: Assign the axes
x-axis = the three conditions. y-axis = mean recall score, starting at zero.
Step 4: Add the finishing touches
Title the graph “Mean recall score by revision condition” and label both axes with units.
Bar graph; conditions on x, mean score on ya histogram here would be wrong — conditions are not a continuous scale
WORKED EXAMPLE
Spot the errors in a graph
A student presents a graph of reaction times. The bars touch, the y-axis starts at 250 milliseconds, there is no axis label on the x-axis, and the title reads “Graph 1”. Identify three problems.
Step 1: Check the bars against the data type
Touching bars means a histogram, which is correct if the x-axis shows reaction time intervals — but wrong if it shows separate conditions. The missing x-axis label makes this impossible to judge.
Step 2: Check the y-axis
Starting at 250 ms rather than zero visually exaggerates the differences between bars.
Step 3: Check the labelling
No x-axis label and units are missing, so the reader cannot tell what is being compared.
Step 4: Check the title
“Graph 1” describes nothing. It should state the variables shown.
Truncated y-axis, missing axis label, uninformative titlethe truncated axis is the one students forget — it is a way of misleading, not just untidiness
💡 Exam tip
Say the word gaps when justifying a bar graph and touch when justifying a histogram. The mark scheme looks for it.
Always start the y-axis at zero unless the question says otherwise, and mention it if you are asked to evaluate someone else’s graph.
Never join the tops of bar graph bars with a line. Only continuous data earns a line.
A graph question usually carries marks for title, axes and units separately from the plotting itself.
Scatterplots are for correlations only. If there is an IV and a DV, it is a bar graph or a line graph.
For the IA, one clear, correctly labelled graph beats three decorative ones.
⚠ Common mix-up
Using a histogram for conditions. Conditions are categories, so they need gaps.
Thinking the gap rule is just style. It signals whether the scale is continuous.
Cutting the y-axis to make a difference look bigger. It is treated as a distortion, not a shortcut.
Plotting raw scores when the mean is what matters. Bar graphs of conditions almost always show means.
Forgetting the units. “Score” is not a unit; “score out of 20” is.
Using a line graph for anything that is not over time. The line implies values existed in between.
Up next: Box Plots and What They Show — a graph that fits five numbers into one picture and tells you the shape of a data set at a glance.
Want this explained one-to-one?
Book a free session with an experienced IB Psychology tutor and get your trickiest topics made simple.
