Just some clarification

[mog123]

Hi,
I was wondering if anyone could help, I'm a newbie to SPSS and I've looked
everywhere for an answer but I can't seem to find it anywhere.

Basically, I am doing a study with two questionnaires measuring two
different variables, they are both likert scales.I will need to do
correlations however with the input of data I've got a few questions.


1) Do I need to make a total for each questionnaires or input the individual
question? and do these need to be on the same dataset?

2) Do I need to make a standardized likert scale because they both vary?

3) Once this is done how will a form a correlation between them?

I know this is quite basic stuff but I've confused myself and need a quick
layman terms answer if possible, thanks if you can help :)



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Just-some-clarification-tp5729376.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Maguin, Eugene]

Not to be mean and skeptical but it's about final project/final exam time and these questions could have originated from those situations. If so, your TA/instructor should be providing guidance. If not, then I'll give answers. Gene Maguin

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of mog123
Sent: Sunday, April 26, 2015 7:17 AM
To: [hidden email]
Subject: Just some clarification

Hi,
I was wondering if anyone could help, I'm a newbie to SPSS and I've looked everywhere for an answer but I can't seem to find it anywhere.

Basically, I am doing a study with two questionnaires measuring two different variables, they are both likert scales.I will need to do correlations however with the input of data I've got a few questions.


1) Do I need to make a total for each questionnaires or input the individual question? and do these need to be on the same dataset?

2) Do I need to make a standardized likert scale because they both vary?

3) Once this is done how will a form a correlation between them?

I know this is quite basic stuff but I've confused myself and need a quick layman terms answer if possible, thanks if you can help :)



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Just-some-clarification-tp5729376.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Art Kendall]

"1) Do I need to make a total for each questionnaires or input the individual
question? and do these need to be on the same dataset?"

How many cases do you have?
How many items on each of the scales?
Do you have complete sets of items for both scales for each respondent?
Have the measurement qualities of the two scales been well established in prior research?
Where is the data now? on Paper? In something like survey Monkey or other computer-aided-administration? In a spreadsheet?
How complicated is the scoring key? Are the items well balanced?

Whether it is more practical to create two data sets and work out the measurement qualities of the scales separately or to do all you work in one data set depends on the particulars of your situation.

"2) Do I need to make a standardized likert scale because they both vary? "
You do not NEED to standardize, your scales, ordinary Pearson correlation in fact standardizes the variables as part of its processing.

However, you might want to also standardize your scale so that you can use scatterplots to help your understanding of correlation.

Unless there are published norms for the scales or unless there is a lot of research using total scores I suggest you use means of the items so that the summative scale corresponds to the response scale.

Again since you are a newbie, you might want to score the scales as means and as totals and correlate the scores using means with the scores using totals and use visualization of the relation.

"Once this is done how will a form a correlation between them? "
when you have the scores use CORRELATION and <Graphs> <Chart builder> to to draft your syntax for scatterplots.

When you have the scatterplots, edit them one at a time in the output file and fi both a linear fit and a loess fit.  Eyeball whether the two fits look pretty much the same.

When you have the output file from teh