What Kind of Test?

#1
Hey guys, I was wondering what is the best type of test for this:

I'm investigating vegan identity. I have calculated an "integration score" for each individual and am seeing how factors affect that.

I want to see how familial support affects the integration score of females only.

Would this be a paired samples test or should I select female cases only and then run an independent samples t-test?? or neither??

Thanks for any help :)
 
Last edited:
#2
integration score - exactly what values are possible for this measure?

familial support - exactly what values are possible for this measure?
 
#3
Hi there, thank you for your reply!
Integration score has been calculated from other questionnaire responses, so is on a continuous scale between 0 and 65.
Familial support is either 0 for support or 1 for no support
 
#4
In that case it sounds like you want an independent-samples t-test with family support as the independent/grouping variable and integration as the dependent/test variable.
 
#5
Hey there, I have run that test, with female only cases selected

I have attached the results for it. I was hoping for some clarification on the results...

the t-result was a minus (-1.593). What does this mean?
I have never come across this before and I can't find it in my text book (Pallant's SPSS Survival).
A quick google left me more confused as I couldn't seem to get the wording for my search right.

This is my written interpretation for the tables:

First, only female cases were selected for, to conduct an independent t-test. This was to compare the integration score of females with and without familial support. There was no significant difference in scores for those with support (M = 11.77, SD = 6.320) and those without (M = 14.81, SD = 5.456); t (40) = -1.593, p = .119, two-tailed.


is this about right? Does it literally only say that there is no statistical significance?
I'm sorry for this, I am very much a beginner at this and definitely have trouble interpreting results, so if you could explain I would very grateful!
 
Last edited:
#6
the t-result was a minus (-1.593). What does this mean?
I have never come across this before and I can't find it in my text book (Pallant's SPSS Survival).
A quick google left me more confused as I couldn't seem to get the wording for my search right.
Short answer: it doesn't matter.

Longer answer: t will be negative half of the time, and whether it is negative on a specific test depends on (1) what statistics program you are using and whether it calculates the difference between the means as mean1-mean2 or as mean2-mean1, and (2) which group in your study you (arbitrarily) decided to call group1 and which you (arbitrarily) decided to call group2. Bottom line, mainly the magnitude of the difference matters, not so much the sign.

This is my written interpretation for the tables:

...

is this about right? Does it literally only say that there is no statistical significance?
Yes, and yes.