Statistics and Shared Variance - need help ASAP (1 Viewer)

[mgg20]

Hello everyone,
I need help ASAP with something. I need help working out a shared variance for my psychology class.

I have all the information provided by my lecturers but I still don't know how to work things out and they are unable to help.
I need to work out the shared variance between two variables (correlation coefficient (r)).

My first variable is Depression (r = -.106, p = .000) and Prosocial behaviour (I don't have the r or p value, but I have the M, SD, min and max).

Second is Stress (r = -.080, p = .000) and Prosocial Behaivour.

Third is stress (r = -.056, p = .063) and Prosocial Behaviour.

My question is, how do I work out the shared variance if I don't know the r and p value for Prosocial behaviour and the only info I have is the M, SD, min and max?

I also need to work out if they are statistically significant, so does that mean that I just need to look at the p values and determine that or do I need to compare them to something?

I'm extremely confused on how to get thee r for prosocial behaviour, as I also need to calculate R2 and the shared variance %.

Thank you everyone!

 

[Eagle Mum]

I can’t help you with the actual answer, but thought that I should post that ‘r’ is the correlation coefficient between two variables, so it’s not an ‘r’ value for each variable, but rather that the ‘r’ value that describes the relationship between depression and prosocial behaviour is -0.106 with p<0.0005.

There’s an error in one of your other two statement - you’ve got two different ‘r’ values for the relationship between stress & prosocial behaviour - I am guessing one of these must apply to the relationship between stress and depression.

 

[mgg20]

I can’t help you with the actual answer, but thought that I should post that ‘r’ is the correlation coefficient between two variables, so it’s not an ‘r’ value for each variable, but rather that the ‘r’ value that describes the relationship between depression and prosocial behaviour is -0.106 with p<0.0005.

There’s an error in one of your other two statement - you’ve got two different ‘r’ values for the relationship between stress & prosocial behaviour - I am guessing one of these must apply to the relationship between stress and depression.

Hi,
I realised that I actually messed up and I do have the r and p value for both variables, so the depression and prosocial is r = -.106 and p = .000 and so on.

I just need to calculate the shared variance now, which I assume is depression x stress, depression x anxiety, and stress x anxiety. Since the r values are negative, I just multiply the two and assume they're negative as well?

 

[Eagle Mum]

Hi,
I realised that I actually messed up and I do have the r and p value for both variables, so the depression and prosocial is r = -.106 and p = .000 and so on.

I just need to calculate the shared variance now, which I assume is depression x stress, depression x anxiety, and stress x anxiety. Since the r values are negative, I just multiply the two and assume they're negative as well?

A negative value means that the relationship is inverse - as one variable increases, the other decreases.

From purely a mathematical approach, I wouldn’t have expected all three values to be negative, since if two variables are inversely proportional to a third, then the first two variables should be directly proportional. However, as I commented in my first post, at least one statement has to be wrong, so the first step would be to recheck all of your relationships and their ‘r’ values before you attempt calculations.

 

[Eagle Mum]

I haven’t performed this specific type of statistical analysis - hoping someone who has studied psychology will jump in, as I don’t have time at present to do any reading.

 

[mgg20]

A negative value means that the relationship is inverse - as one variable increases, the other decreases.

From purely a mathematical approach, I wouldn’t have expected all three values to be negative, since if two variables are inversely proportional to a third, then the first two variables should be directly proportional. However, as I commented in my first post, at least one statement has to be wrong, so the first step would be to recheck all of your relationships and their ‘r’ values before you attempt calculations.

To be honest, this is all extremely confusing since I haven't done math in a really long time and this is way more advanced than what I'm used to. Essentially all the information was given to us by the lecturer so I'm only using the information that was given to me. I'll insert a photo just In case I'm totally reading it wrong.

 

[Eagle Mum]

To be honest, this is all extremely confusing since I haven't done math in a really long time and this is way more advanced than what I'm used to. Essentially all the information was given to us by the lecturer so I'm only using the information that was given to me. I'll insert a photo just In case I'm totally reading it wrong.
View attachment 31851

This does make more sense - all of these negative symptoms are inversely correlated with prosocial behaviours and are likely to be directly correlated to each other (eg. stress & anxiety often occur together).

Edited: Whilst ‘r’ the correlation coefficient has a sign to indicate the relationship between variables, shared variance is looking at size of effects/correlation and is based on r^2.

Multiple regression analyses is used to determine the shared variance of factors (independent variables) that contribute to an outcome (dependent variable), whereas there may be significant correlation between the independent variables that is independent of the third (consider the Venn diagram where two circles overlap with and without overlap with a third circle). I don’t know the details of your specific exercise but this site helps to explain the concepts & calculation steps for multiple regression analyses, including how to take into account the various components of overlap: http://faculty.cas.usf.edu/mbrannick/regression/Part3/ImportanceNarrative.html

Look particularly at Tables 5.1 & 5.2 and Figures 5.1- 5.4.

 

[Eagle Mum]

Hi @mgg20, how did you go?