Hello everyone!
I've got four variables measured by different questions in an inventory. I have inputed them in spss so that I can get frequencies of each question. Now that this is done, I don't know how to sum all the questions so that I can get a value for each variable. To make things less easier there is the fact that the questions aren't valued in the same way but, in some cases, the same answer can either add a point to the final sum or not.
How can I accomplish this with SPSS?
What I have tried is to revalue every single question according to the invetory's criteria, and then I have sumed all those new "variables". Sadly, that didn't work because the values didn't match with the ones I have got with a different software.
Any suggestions?
I am attaching SPSS file which is in Spanish.
Thank you so much!
Hi!
Can you be a bit more precise: why do you want to sum them? Should these 4 variables be one factor? If yes - go to alayze-dimension reduction-factor analysis and build one factor from them.
Or maybe you just imputed them incorrectly? You should have 4 variables (in the variable view),an all your participants as a column in the data view. One participant - 1 value for each variable. In this case, you should reenter your data into SPSS.
Or if you want to see the correlations among these variables, you go to analyze-correleations-bivariate.
Best,
Eugene
I can't understand why I can not upload the file. Guess RG was not made to let this kind of files to be uploaded.
What I tried to explain, with not much success is:
1. I have four variables (Personal, Familiar, Educational, Social)
2. Each variable is tested in the same inventory throughout 123 questions. Each variable have like 30 questions.
3. In my variables view, I see all those questions and also some sociodemographic data not really relevant for the point of my question.
4. In my data view, I see all the participants, which by the way are 213 students.
5. There are three possible answers for each question: Yes, No, and ?
6. For SPSS, I have registered the answers like follows: Yes = 1; No = 0; and ? = 99
7. Each variable is measured trough more or less 30 question scattered all along the questionnaire.
8. When it comes to calculate the variable, if you answer Yes then you get 1 point to your final score in that variable; if you answer No or ? you get no point at all. But this is only the case of some questions. In some of them, if you answer No you get the point; and if you answered Yes you don't.
So...
How must I proceed to get the total sum for each variable (Personal, Social, Familiar, Educational)?
I have tried making of each question a new variable (for SPSS, not for the test in itself) and to this new variable I assigned a new value, not according to the previous system I created before (Yes=1; No=2, ?=99) but according to the way the test is calculated... that is, I give 1 when "Yes" means to add one point, and I give also 1 to all those questions that "No" means to add one point. (Taking into consideration that the other two possible answers get 0 points if chosen). This, however, didn't work.
So... what do you think?
You need to build super variables. It is done in SPSS through the dimension reduction procedure. Go to Analyze - Dimension reduction - factor. Choose principal component analyzes, choose restricted number of factors, restrict by 1. Add all items from one factor. Options - click on save as regression. Press ok. Now you have a single variable which is build from all items. It is quite complex procedure if done correctly. You have to check convergent and discriminant validity, internal consistency. And exchange all 99 by Missing values, otherwise the Programm will treat them as ratios.
Look for tutorials on YouTube. If still problematic, I can help you over Skype or TeamViewer on Wednesday or Thursday.
Best,
Eigene
Do you have any specific YouTube video that you consider can help me out?
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS