SPSS question

[psych844]

Hello,

I just want to add some error bars to my line graphs, after having ALREADY produced the graph itself. I tried clicking on the individual graphs to make the addition, and while adding error bars option does exist, it is not clickable!

Cheers

---

Members don't see this ad.

 

[Harry3990]

Not helpful for your question, but I'd recommend using excel for graph creation. Looks better and way more flexible.


Sent from my iPhone using SDN mobile app

 

[MCParent]

Hello,

I just want to add some error bars to my line graphs, after having ALREADY produced the graph itself. I tried clicking on the individual graphs to make the addition, and while adding error bars option does exist, it is not clickable!

Cheers

It's easier to do it in excel; you can easily plot asymmetrical 95% CIs in there.

 

[psych844]

It's easier to do it in excel; you can easily plot asymmetrical 95% CIs in there.

Am I able to take my SPSS Graph, paste it into excel, and then somehow add error bars? How would I do that? Cheers

 

[Harry3990]

I don't think so. you would need to paste the descriptive stats table. Then recreate the line graph in excel and select the correct numbers for the data.


Sent from my iPhone using SDN mobile app

 

[psych844]

Strange that you can't just add error bars after the fact on SPSS.

 

[Ollie123]

SPSS graphs are really quite limited. I wouldn't be surprised if it has to do with it being a line graph vs. bar graph - I think its possible but you may need to go to syntax to do it.

That said, agree with others that SPSS sucks for graphing. SAS is just as bad unless you have one of their fancy add-ins. Excel or the myriad of other options are far better in most cases. SigmaPlot, TeraPlot or even R or Matlab (as long as you don't mind syntax) are all great if you have access.

 

[bcliff]

If you're using syntax, you would need to figure out the right code and then rerun the command. Never thought about pasting descriptive stats tables from SPSS into excel for the sake of graphics -- Good idea!

 

[MCParent]

You write a matrix that's like:
meanG1T1, meanG2T1, meanG1T2, meanG2T2
upperboundCIG1T1, upperboundCIG2T1, upperboundG1T2, upperboundG2T2
lowerboundCIG2T1, etc etc etc
blah blah blah, or whatever makes sense for your data. Then in the "error bars" option of excel you can make select the fields for errors and can make asymmetric error bars.
I have a ppt I made on how to do this but I can't upload files on here

 

[Therapist4Chnge]

Can you email me the ppt?

 

[MCParent]

done