Box Plot and Anova Questions

[PerMagna]

Hi there,

first of all. Thank you for that amazing piece of software.

I have three questions regarding sofa.

1.) i have highly significant results (independent t-test). then i created a boxplot.nonetheless at the boxplot there is no star indicating significance.
how can i add a star? i know how to do it with gimp (or within SPSS), but is there an option within sofa? i see a small star at the symbol of "boxplot". where is it! :D

2.) i have other data (3 groups, each group n=5, intervall, normal). i used anova to check significance.
i always get this error message at p abnormal7  "Unable to calculate overall p for normality test"
why? i checked data for normality. better do a kruskal?

3.) my calculated p relates to which of those 3 groups?  Blue, Green, Yellow.
i set from group "blue" to group "yellow".
Results of ANOVA test of average Hugs for Color groups from "Blue" to "Yello"
It calculates the mean of group "green" aswell.
i have significant results, does that mean that blue is more likely to get more hugs then green and yellow?
or does it just say that there is a difference within the three groups but further test are required to check which one?

thanks i advance!

[PerMagna]

3.) ok i realized i need a post-hoc test. are there post-hoc tests in sofa?

[Grant Paton-Simpson]

Hi,

On 19/02/16 23:40, 'PerMagna' via sofastatistics wrote:

Hi there,

first of all. Thank you for that amazing piece of software.

Thanks - glad you like it.

I have three questions regarding sofa.

1.) i have highly significant results (independent t-test). then i created a boxplot.nonetheless at the boxplot there is no star indicating significance.
how can i add a star? i know how to do it with gimp (or within SPSS), but is there an option within sofa? i see a small star at the symbol of "boxplot". where is it! :D

Interesting - how exactly would that work - any examples etc?

2.) i have other data (3 groups, each group n=5, intervall, normal). i used anova to check significance.
i always get this error message at p abnormal7  "Unable to calculate overall p for normality test"
why?

If the data isn't sensitive you can send me a copy at gr...@sofastatistics.com and I'll try to replicate the error.

i checked data for normality. better do a kruskal?

3.) my calculated p relates to which of those 3 groups?  Blue, Green, Yellow.
i set from group "blue" to group "yellow".
Results of ANOVA test of average Hugs for Color groups from "Blue" to "Yello"
It calculates the mean of group "green" aswell.
i have significant results, does that mean that blue is more likely to get more hugs then green and yellow?
or does it just say that there is a difference within the three groups but further test are required to check which one?

The latter. SOFA doesn't currently have the required test but it's starting to feel like a real gap.

All the best,
Grant

thanks i advance!

--

---
You received this message because you are subscribed to the Google Groups "sofastatistics" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sofastatistic...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.