Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Adding error bars to bar chart
From
Nick Cox <[email protected]>
To
"[email protected]" <[email protected]>
Subject
Re: st: Adding error bars to bar chart
Date
Tue, 4 Mar 2014 17:25:58 +0000
I don'f think anyone quite recommends that as the method of choice,
but it is one possibility.
Much depends on
generally, quite what you want or need to show
specifically, whether your SEs are necessarily calculated separately
or are just SD/sqrt(sample size) and as such can be calculated by some
existing command
and whether you +/- SE or some confidence interval.
Nick
[email protected]
On 4 March 2014 16:39, Chris Yang <[email protected]> wrote:
> Thanks for all the suggestions and references.
>
> From the thread mentioned by Nick, it seems that it's better just to
> use the error bar chart (-serrbar-), and possibly include the raw
> data?
>
> On Tue, Mar 4, 2014 at 9:28 AM, Nick Cox <[email protected]> wrote:
>> The briefest suggestion is "Don't do that" for reasons developed
>> (e.g.) in the thread starting
>>
>> http://www.stata.com/statalist/archive/2013-05/msg00293.html
>>
>> and its references.
>>
>> A less cryptic comment would be that this almost always (1)
>> over-reduces the data (2) places most emphasis where it is least
>> appropriate (3) is more difficult to interpret than easy alternatives.
>>
>> There is no difficulty in principle, except that if you think the
>> place to start is -graph bar- that is quite wrong. You would need to
>> start with -twoway bar- and then add other -twoway- elements. By the
>> same token, you would need to calculate everything you need as
>> variables before you enter -twoway-.
>>
>> Nick
>> [email protected]
>>
>> On 4 March 2014 15:09, Chris Yang <[email protected]> wrote:
>>
>>> Is there a way to add SEM error bars to a bar chart? I have the SEM
>>> calculated for each bar in the data, but drawing it on top of the bar
>>> seems not so straightforward in Stata.
>>>
>>> Any suggestion?
>> *
>> * For searches and help try:
>> * http://www.stata.com/help.cgi?search
>> * http://www.stata.com/support/faqs/resources/statalist-faq/
>> * http://www.ats.ucla.edu/stat/stata/
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/faqs/resources/statalist-faq/
> * http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/