Compute a mean score from several variables in SPSS
HTML-код
- Опубликовано: 17 сен 2015
- If you have 2 or more variables measuring the same factor or construct (often this happens in questionnaires), then you can combine them into one measure using a mean score or total score. This will create a new variable in your dataset. This new variable is what you will use for further analysis.
This video shows how to compute a mean score.
oh my! I am dyslexic and my lecturer put step by step guides using words. Without your tutorials I swear I would have failed my module! THANK YOU THANK YOU!!!
This was so great, thank you so much! I didn't even know you could specify how many variables to add in the mean, I gained more knowledge than I intended :)
I'm deeply grateful for this tutorial.
You are just the best. Thank you for the tutorials
Your videos have been very helpful for me. Thank you so much & have a great day. :-)
The video has been very helpful. Grateful
Omg thank you! I was trying to write the code in SPSS and I kept getting this stupid error code but when I did it this way...boom no error! You saved me!
short and clear, Thank you so much. one thing, after mean is done, how to analyze that mean?
Thanks you! Super clear. Do you also have a video on computing two variables with different metrics?
Hi, I am currently doing a survey analysis. I used 4 questions to represent each of my independent and dependent variable. So I have 4 set of responses for each variable in the study. In order to further analyze them for instance to do a regression analysis between my dependent and independent variable, do I need to compute all 4 responses that related to my variable as shown in this video?
great video, thanks
Thanks a lot. I would like to use a single composite variable of a 4 items likert-scale variables, to make corrolation in regression. i would like to know if i should use the MEAN or SUM of the all the indicators i put together to form one composite variable for my corrolation and regression analysis. your reply will help a lot.
Is it appropriate to calculate correlation on lickert scale data....
If you have say 6 variables (value 0-100) for one indicator/concept, only 2 were answered and I can't give a complete picture without all variables; would it make sense to calculate the average of the two varaibles, divide it by 6 and multiply by 2? The purpose would be to consider that only 2 variables shouldn't be allowed to give a final score of 100 if say both variables were measured to be 100 (making the average 100).
please help me to give some information that how i can marge three types of value (ordinal nominal and scale ) in same scale value.
Good and brief Solution
Thank you that was very practical to do and well presented
helloo... can we calculate correlation for licker scale...
kindly guide
Yes, both variables should be on a continuous or an interval variables.
I dont know what I m doing wrong. I followed the steps to add variables together to create a new varaible. I can see the syntax is correct, but the new variable is not appearing in data or variable view for me to be able to use it in further manipulation in Spss. Please, help me, I need urgent assistance.
THANK YOU!
Love your video. Thank you so much for sharing! So why would you calculate 4 different means? Is it necessary in a real analysis world?
So your means assume the difference between undecided and agree is the same as the difference between agree and strongly agree? You are turning an ordinal variable into an interval/ratio variable (can also see that from your variable view window). I don't understand the point of that unless you want to (incorrectly) use parametric tests.
The "OK"button isnt available for me, after I followed your steps
it says "Incorrect variable names :either the name is more than 64 character or...."
did you find out the reason for this error
you need to recode the data because it can't read data in numeric
Make sure the measure of the variables you're using is set to i.e. ordinal in Variable View in SPSS
Hello, i have got 3.5 value of one variable and it is neither making it agree nor neutral. can you please help ?
1= strongly dis, 2 dis 3 neutral, 4 agree, and 5 strongly agree.
This is because you calculated the mean you cant be half strongly agree. Instead use the median.
thanks
Dear dr. Can you suggest the compute procedure, if we want to choose the 4 variable (numeric), which one is the highest, and thus detect those values? Detect means choose the name of variable to be appear
How to calculate mean scores
1:57 rein
piste ka
patakag tuldo