IA Graph: How to create the perfect graph for your IB Psychology IA

Travis Dixon Internal Assessment (IB)

There's no excuse for submitting a poor graph for the IA. It's really easy.

Your graph must be accurate and “address the hypothesis.” In this post we’ll look at how you can create the perfect IA graph for the IB Psychology IA. 

Your research hypothesis will be based on the effect of an independent variable on a dependent variable. To test this hypothesis you created two conditions (I hope) in your experiment to see the effects of the IV on the DV. Therefore, your graph should clearly display the results from the two conditions.

Read more:

Column Charts

Most IAs will use a column chart. Why? Because you want to visually compare two conditions of an experiment. A column chart is most often the easiest way to do this.

Some students make the mistake of having one graph per condition. Do not do this. Remember your using the graph so your reader can see a visual comparison of the conditions of your experiment. Separating this into two graphs defeats the purpose.


Histograms: A histogram is probably not appropriate for your IA. Histograms display a range of continuous data, whereas you are comparing to distinct conditions.

When not to use a column chart: I can only think of one example when a line graph would be appropriate. This would be if you were doing a study on the primacy and recency effect and you wanted to show the serial position curve. However, in my advice on doing Glanzer and Cunitz I advise to study either the primacy or recency. This means you should also have a column graph that allows a simple comparison.

Post a comment if you have an example of a study with results best displayed in a graph other than a column chart.

What not to graph

Do not attempt to graph your spread (dispersion) in the columns. e.g. your standard deviation, range, etc. Your hypothesis has nothing to do with how spread out your data is, so you do not need to graph this. Some students choose to add error bars to represent their standard deviation. This is fine.

Do not graph your raw data, either. We are looking at comparisons across conditions and showing raw data does not help us do this.

The Perfect IA Graph


  • Clear title (Tells me what the graph is showing)
  • y and x axis labelled
    • Y axis: what was measured (the dependant variable)
    • X axis: the two conditions (the independent variable)
  • Graph presents the average (mean, median or mode) of the two conditions and not raw data
  • y axis is not misleading (e.g. it includes the range of possible values and begins at 0)
  • It is clear which condition is which

Spot the Mistakes

Imagine you’re an IA examiner. A student has done Loftus and Palmer (1974) for their IA. They have created the following graph. What feedback would you give them for their final draft?