## Compute Variables and Analyses

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

MonkeyStudent
Posts: 2
Joined: Tue Jan 03, 2017 10:41 am

### Compute Variables and Analyses

Hey,

I'll try to explain it step by step. In my study, i have asked on a 10-point scale what the respondents think about the wrongfulness of 11 different crimes for which i have 11 different variables in SPSS. I need to make a scale so i have computed them (MEAN and SUM). I want to see if there is a difference between two groups for which i have chosen the Independent t-test.

Question 1: Should i use the computed MEAN or SUM scores of the variables in an independent t-test? Which give a more clear (better) output?

Question 2:Same question for when using (single and multiple) regression analyses

statman
Posts: 2756
Joined: Tue Jun 12, 2007 12:08 pm
Location: Florida, USA

### Re: Compute Variables and Analyses

If understand, load in the raw Likert date but make sure you define the variables as Ordinal

Also .............
See the note below

Statman
Statistical Services
MonkeyStudent
Posts: 2
Joined: Tue Jan 03, 2017 10:41 am

### Re: Compute Variables and Analyses

My guidance counciller said i had to make scales to avoid Type I errors, because i have allot of variables. I need to know if there is a difference between Gamers and NonGamers for their perception towards the wrongfulness of crimes.
He is on holiday vacation now for another week, so i was wondering if he meant by 'scaling', that i need to compute my variables. If so, my original question is: what output should i interpret, SUM or MEAN scores of the computed variable to present the data in my paper.

I use a laptop with Windows 10, 64bit; SPSS Version 24

What i have done is the following:
Raw data/Variables
Var3.2.1 = Breaking in a home without violence
- Values: 1=1, 2=2, 3=3,... 10=10, 999=not filled in
- Measurement level: Ordinal
Var3.2.2 = Stealing with violence
- Values: 1=1,...
- Measurement level: Ordinal
....
Var3.2.11=Discrimination
- Values: 1=1,...
- Measurement level: Ordinal

Compute
SUM (Var3.2.1,Var3.2.2,Var3.2.3,Var3.2.4,Var3.2.5,Var3.2.6,Var3.2.7,Var3.2.8,Var3.2.9,Var3.2.10,Var3.2.11)
- Variable name= Wrongfulness_SUM
- Measurement level: Scale

MEAN (Var3.2.1,Var3.2.2,Var3.2.3,Var3.2.4,Var3.2.5,Var3.2.6,Var3.2.7,Var3.2.8,Var3.2.9,Var3.2.10,Var3.2.11)
- Variable name= Wrongfulness_MEAN
- Measurement level: Scale

Test for analysis
Independent T-test
- Test Variables
==> Wrongfulness_MEAN
-Measurement level: Scale
==> Wrongfulness_SUM
- Measurement level: Scale
- Groupung Variable
==> Not_Gamers (NG; defined as '0')
- Measurement level: Nominal
==> Gamers (G; defined as '1')
- Measurement level: Nominal

Output
Wrongfulness_SUM & Wrongfulness_MEAN:
http://imgur.com/a/y6ByZ#WyozJc7

Apologies for the imgur link (lacking skills in img posting on forum)

Question 1: Is this the right proces for scaling 11 variables to 1 variable?
Question 2: What output data should i interpret?

### Who is online

Users browsing this forum: No registered users and 5 guests