# Which statistical analyses do I need to use?

#### Wittingly

##### New Member
Hello everyone! :wave:

I'm having some trouble resolving two assignments with SPSS.

First one is that I need to analyse if there's a difference between poor and rich European countries (using a GDP per capita index variable) and link it with thinking that the environment is the most important issue.

I recoded this environment variable (before it was "what is the most important issue for your country?": economics, crime, health ...) so before I recoded it, it had 15 categories or so. So I decided to recode it in 2 categories those who thought that environment was the most important and those who didn't. There's also a Likert scale variable that I could use instead of this 'environment' variable, but I think it would be too difficult.

The GDP per capita variable is a ratio variable going from 44-..., while the environment variable is a nominal variable with two categories (environment is important and environment isn't important).

GDP_Capita_Index
Frequency Percent Valid Percent Cumulative Percent
Valid 44,00 1.003 6,1 6,1 6,1
51,00 1.000 6,1 6,1 12,2
74,00 1.159 7,1 7,1 19,3
80,00 1.428 8,7 8,7 27,9
85,00 1.082 6,6 6,6 34,5
100,00 2.560 15,6 15,6 50,1
112,00 928 5,6 5,6 55,8
115,00 1.211 7,4 7,4 63,1
118,00 1.407 8,6 8,6 71,7
119,00 1.142 7,0 7,0 78,7
123,00 1.181 7,2 7,2 85,9
126,00 1.019 6,2 6,2 92,1
127,00 1.305 7,9 7,9 100,0
Total 16.425 100,0 100,0

Here's the original .sav file, if my explanation isn't clear: https://docs.google.com/file/d/0BwhY_BiXpRQ7TnZjUHlGRlBtM28/edit

So what I need to do, is to find if there's a difference between poorer countries and richer countries and thinking environment is the most important issue? I was thinking of using an independent sample t-test...

Second question, I need to take this "environment" variable and use it on different variables like gender, being member of a union, place of living, age group, placing on left-right scale. Making sure I only use "pure effects" (probably main effects) effects that are not caused by other variables. I think it's highly likely I'll need to use the GLM model, because that's the only model I know of, where you can check for main effects.

I would greatly appreciate if one of you could set me on the right path, you would have my eternal gratitude! :yup:

Last edited: