# Why are error bars used instead of error blobs or error rectangles?

In reporting experimental results, it is common to use error bars to communicate uncertainty in measurement of the quantity represented by the y-axis.

But, usually, there would also be uncertainty in measurement of the quantity represented by the x-axis.

However, I have rarely seen that being shown on a graph throughout academia and research community.

Why is it so? Is there any merit in showing error rectangles instead of error bars?

It’s perfectly possible to use error bars on the x axis as well as the y, so the error bars look like a cross. Rectangles or blobs aren’t normally used because they would obscure the point in the plot corresponding to the mean value.

• Thank you! Could you also throw some light on why that isn't the norm? – Ritesh Singh Dec 15 '19 at 18:09
• I’m not an experimentalist, but it is probably because in many cases the value on the $x$ axis is known to such a high precision that its error bars would not be visible. – DavidH Dec 16 '19 at 1:45