Comparing means between sample and subsample

#1
I would like to compare the mean score on personality questionnaire between my total sample (n=97) and my subsample (n=44) to see if there is a significant difference. I have absolutely no idea which statistical test I can use for this, since I mostly use independent or dependent sample t-test and both of them don't seem to fit here.

Can anyone help me? You would be a lifesaver!
 

Karabiner

TS Contributor
#2
it does not make sense to compare subjects to themselves. Imagine your subsample were n=90, would you then expect that there's a meaningful way to perform a statistical test to compare them to the total sample n=97 ? You can just compare n=44 to the rest of the sample (n=53). Wouldn't that give you all the answers that you are looking for?


With kind regards

K.
 

hlsmith

Less is more. Stay pure. Stay poor.
#3
I have seen journal referees say to compare descriptives of non-respondents to the full population of eligible respondents to show they are comparable and I guess just missing at random. Though, I think a better way to do it is develop a logistic model with the outcome, missing (Y/N).


An-KatrienDR,

Can you tell us why you want to do this, so we can better understand the appropriateness of options.
 
#4
I did two experiments (aversive conditioning tasks), one with the total sample, and one with a subsample, so I want to see if the mean scores on a personality questionnaire don't differ significantly from each other (between the total sample and the subsample).
 

hlsmith

Less is more. Stay pure. Stay poor.
#7
There are mental red flags popping up every time you post more information. How were the individual that get two interventions selected?