1
$\begingroup$

I am seeking some general ideas on how to more effectively present two sets of confidence intervals in a scatterplot. There are a total of 10 point estimates and the associated 20 CIs, 10 on each dimension. The difficulty is that, some confidence interval is very wide, so if I draw them vertically and horizontally for each point, then the plot would be very ugly... Is there a more effective way to incorporate both CIs for each point in a scatterplot? It may not be limited to drawing lines, other solutions using different visualization methods are also good. Thank you.

$\endgroup$
1
  • $\begingroup$ Please post the "ugly" version of your visualization so that we can see the problem more clearly. $\endgroup$
    – Kostia
    Jul 2 '16 at 8:09
2
$\begingroup$

Without seeing the data, I would suggest to use log scale for the dimension with wide intervals. If this simple trick does not solve the problem, please add your current visualization to your question.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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