Bimodal vs Unimodal

Hi there

I have a question about Question 3 b)ii from the 2018 paper

So the marking schedule says that the distribution of energy consumption in the North Island is roughly unimodal and symmetrical. However, in my answers, I said that North Island follows a bimodal distribution with two peaks at around 650 and 750.

I know that bimodal data has two peaks, but I often get confused when it is appropriate to label graphs as unimodal and when to label them as bimodal, especially when the data is clustered quite closely. Would I be incorrect in this case?


Bimodal distribution is often a result of graphing data from two distributions. They have two separate clusters with a drop to the minimum between them. I can see how you came to this conclusion from the graph and it would be up to a marker to accept your description or not. Sometimes due to sample size there can be more than one interpretation of results. If you can’t decide between bimodal and unimodal distribution it is probably unimodal, but in your report you may mention both distributions.

Hi again @11111111

Yes I agree with @Freydis, for this example it is a small sample size so isn’t clear. It is important you back up your interpretation with why you think it is that and include some values. There is often more than one way to see the graphs!