# Thread: Standard error bar question.

1. ## Standard error bar question.

Im doing a lab on mitosis and have calculated the SE already for all the categories. I am almost certain that when the instructions state (± 2 SE) for the error bars that you multiply the SE by 2 and use that value for ur error bars. Most of my group members don't agree with me.

2. ## Re: Standard error bar question.

Agreed with the +/- 2*SE, which would be your 95% confidence interval. Just call them that and the reviewer will understand without abiguity.

3. ## Re: Standard error bar question.

I agree, so I would multiply the SE by two and use those values for my error bars

4. ## Re: Standard error bar question.

I didn't stare too hard at your attachments, but if you are looking at means - YES. The bars would represent your 95% confidence intervals for each mean. Of note, sometimes these whiskers mean different things based on the procedure used, so it is always nice to provide a legend describing what your whiskers represent.

5. ## Re: Standard error bar question.

Originally Posted by Rsahu1919
I agree, so I would multiply the SE by two and use those values for my error bars
I would probably look up the critical t-values to multiply the SE by for your case. Unless you have a very large degrees of freedom, the appropriate t-value will be meaningfully larger in magnitude than 2 for a 95% level of confidence. On top of that, 2 is already a "rough estimate" of the "actual" +/-1.96 that's rounded from the z-distribution. Essentially, you're "approximating" the 1.96 with a 2, and that's assuming you have enough degrees of freedom to justify going off the z-distribution (which, judging from that picture, your largest group seems to be around 10 or so, and you probably can't get away with using the z-distribution).

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts