nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

What This Mean Diamond Lines (before the edges) mean?

WebDesignesCrow

Hello expert,

I would like to ask what these 2 green lines in the mean diamond means?

I know the edge supposedly to represent the 95% of confidence means (upper & lower) and the centre green line means the avg.

How does JMP calculate the 2 green lines in the mean diamond?

Thank you

Accepted answers

ruthhummel

The endpoints of the diamond are the CI for the mean.

The lines near the top and bottom of the diamond (called the "overlap marks") are a fraction (sqrt(2)/2) of the CI, calculated by group mean ± (half CI)*sqrt(2)/2. These marks are only relevant if you have two or more groups to compare and they have exactly the same sample sizes in each group. In that very specific circumstance, these lines show if two group means are statistically significantly different by their overlap with each other. In the picture below, Group B is not statistically significantly different than Group C because they overlap in their "overlap mark" regions. On the other hand, Groups A and B are statistically significantly different, because they do not overlap in those regions.

All comments

txnelson

The lines are called Overlay Marks. Here is their definition, copied directly from the JMP Help page for Mean Diamonds

"The lines near the top and bottom of the diamonds are overlap marks. For groups with equal sample sizes, diamonds that overlap within the overlap marks indicate that the two group means are not significantly different at the given confidence level. Overlap marks are computed as group mean"

WebDesignesCrow

Thanks. I've found the explanation as well for SAS.JMPStatistics guide (2007) but still confused.

Basically, when I calculate 95% confidence interval using formula, CI = Mean +/- 1.96*SE with SE = std/sqrt(N), the values are the same with overlap marks.

Then, how do we get the edges?

Data: 0.13, 0.12, 0.12, 0.11, 0.1

txnelson

I have not found the answer for that. Another Community member may have the answer.

ruthhummel

The endpoints of the diamond are the CI for the mean.

WebDesignesCrow

Thank you @ruthhummel for the explanation.

If I have only 1 group in the chart, is there any way to turn-off the overlap marks to avoid confusion?