10 hints to make the most of teaching and academic conferences
7 December 2016
Has the Numeracy Project failed?
12 December 2016
Show all

A student asked me on our Facebook page to help with an assignment. It got me thinking again about the nature of answers in statistics, and the challenge of communicating through graphs.
The student gave no explanation, but rather a set of data and a choice of graph-type. I have changed the data a little. As the data was about numbers of men and women in different roles in an organisation, I assume the question regards gender equity.
I have used the ubiquitous Microsoft Excel to create some candidate graphs, and I would like you to vote on which you think is the most useful. (And yes – I know that it depends, but in the absence of further information, I’m interested to see what the crowd comes up with.) Pie chart was not an option. Pie chart is almost never an option.
Here are four possible answers for the student. If you click on the graphic it should get bigger and clearer.

Which is the best graph to use?

Which is the best graph to use?

I will post later with my opinion, but I would rather get yours first.


  1. Alex says:

    What I really want is a modified version of B with a reference line at 50% and with the width of the bar proportional to the number of people in the role.

  2. Penny Robinson says:

    I’d like a graph that is a mixture of Graph A and Graph B.
    I.e. Graph A, but with “% in role” on the y-axis instead of “Number in role”.
    Reason: The total number of people in each role differs – more “other”, less “finance” and the others have similar numbers. To compare the % of males and females across the roles, we need % not number. And I like a bar chart instead of a stacked bar chart.

  3. […] Please comment on the article here: Learn and Teach Statistics and Operations Research […]

  4. John Q Public says:

    “Best to use” for what? What is the essential story one wishes to tell using these data?

  5. […] Late last year I posted four graphs of the same data and asked for people’s opinions. You can link back to the post here and see the responses: Which Graph to Use. […]

Leave a Reply

Your email address will not be published. Required fields are marked *