how to enter certain data, and should i use t test?

#1
Hi all,:wave:
I'm new here and need life saving help with my MEd dissertation! I'm studying moral development (SRMS SCORE) and play opportunities (TIME SCORE) and the differences between urban and rural areas.
My variables are as follows: age, ***, region, SRMS moral dev score, and time playing with social interaction, time for play without social interaction.

I have the data for each participant (300) entered in SPSS, except for the time variables... these I only have the average for the 2 regions; urban and rural.

I was thinking that I might run a t test? Is this correct?
I don't know how to enter the data for 'time' variables as I only have the average, and not individual, unlike all my other variables... can anyone advise me on how to rectify this or work with this?

I really appreciate any help that you can give... I'm at my wits end!!

Thank you, T. :)
 
#2
Yes, you could run a t-test to determine if there is a difference between the moral development scores in 2 distinct regions (urban vs. rural). Then if the difference is significant (probably use p=.05), you can start looking into why.

For the time played, you can still run t-tests without the individual points... if you have the standard deviations and the sample size. Without the standard deviations, it'd be hard to tell whether the between group variance is bigger than the within group variance. And without the sample size, it'd be hard to tell how relevant your results are. As for SPSS, I have no experience in it. So, to summarize: yes, I suggest t-tests to start. Then you can look into ANOVA or multiple regression to see why the two data sets are different.
 
#3
If it was me, I would start by going back to my hypothesis - what exact questions do you want to ask as this will influence which statistical test you use. If you want to know if there is a difference between regions on moral dev score, yes a t-test would be appropriate. If you have reason to think that age and gender might be different between the groups as well then do an anova. Are all your participants from the same school year or not? do you want to know if scores are different between year groups etc.

As for time score - how was this data collected? If you have, for example, averages for different schools within each region or different classes then you will have some variance within groups, which means you can do an t-test (although non-parametric might be more appropriate). If you just enter the same score for each participant within each group i.e. if the average for urban group time with interaction is 40 then give everyone in that group a score of 40, then SPSS won't let you do a t-test on it (well, it will but it will give you t values with massive exponentials and no variance).

You could always use an arbitrary cut off point for time and recode it into a dichotomous variable i.e. a time score above 40 is "lots" so give a score of 1, and a time score of less then 40 is "non lots" so give a score of 0. This might be a better way of looking at it, but again depends on how you got the data and what you're asking i.e. is time one variable with two levels or do you want two different variables?

Hope this helps!