[Question] Is it appropriate to use standard error as error bars when visually comparing means?
Hi everyone! I'm just getting started in the field of biostatistics, so apologies if this is a basic question.
I'm working with a bar plot comparing means between different groups, and I'm wondering whether it is appropriate to use the standard error (SE) as the error bars.
My main goal at this stage is to get a visual sense of whether there might be significant differences between the groups. We will later perform more formal statistical tests to evaluate those differences, but I'm unsure about the best way to represent variability or uncertainty in the plot.
Is using mean ± SE appropriate for this purpose? Can the overlap (or lack of overlap) of SE error bars tell us anything meaningful about statistical significance, or would confidence intervals be more appropriate?
Thanks!
