This is a question about data visualisation - I am sorry that it is probably quite elementary and possibly only just relevant to CrossValidated.

I would like to create a figure that will help people visualise some tabulated data, which shows point estimates for cost and effectiveness together with confidence intervals. As I have envisaged the figure, there would be four quadrants with the centre of the X representing no cost but also no effect. I have created something in MS Paint (!) to illustrate what I am trying to achieve but would like to do this quickly using tabulated data and produce something that could be used in a publicly-facing report.

I'm sure that I've seen similar figures around before and wondered whether anyone could tell me (a) what they're called and (b) what my options are in terms of creating one myself?

I have some (not much but probably enough to work it out once I am heading in the right direction) experience of Excel, Stata, R, and Tableau.


  • 2
    $\begingroup$ What is the meaning of the intervals? As a comment: not every plot has it's own name, so there doesn't have to be any name for it. $\endgroup$
    – Tim
    Commented Aug 13, 2018 at 9:43
  • $\begingroup$ The intervals around each point estimate are 95% confidence intervals in my dataset. My data is just organised in Excel, i.e. columns "Drug", "Effectiveness", "Effectiveness lower 95% CI", "Effectiveness higher 95% CI", "Cost", "Cost lower 95% CI", "Cost higher 95% CI". $\endgroup$ Commented Aug 13, 2018 at 10:15
  • $\begingroup$ So you should flip the axes $\endgroup$
    – Tim
    Commented Aug 13, 2018 at 10:25
  • $\begingroup$ Thanks. Do you have any thoughts as to how I could create this without drawing lines in MS Paint? ;-) $\endgroup$ Commented Aug 13, 2018 at 11:05
  • 1
    $\begingroup$ It seems to be a scatterplot with added confidence intervals, so it should be up to few lines of code in R or Python, surely this is possible in other software, but asking about the code would be off-topic in here, so you should rather refer to support forum of your favorite software. $\endgroup$
    – Tim
    Commented Aug 13, 2018 at 11:23

1 Answer 1


I've seen scatter plots with confidence regions (for example, with ggplot's stat_elipse in r), but typically all the observations for each group are plotted, not just a single point that represents the group mean.

Since you only have the points and CI bounds, you could just plot the points with vertical and horizontal lines to indicate the CI. This page describes how to do it in Excel https://peltiertech.com/custom-error-bars-in-excel-charts/


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.