Back
to top
SPSS - create histogram of Pearson correlations
Watch
Announcements
Page 1 of 1
Go to first unread
Skip to page:
Hello,
I need to make a histogram with the Pearson correlations I calculated with SPSS to see how the correlations are distributed. When I click on the output table I am somewhat able to make one, but this includes also the correlation of Variable1 with Variable1 (=1) which I of course do not need.
Is there any other possibility? Can I make a new variable with the Pearson correlations for the datas? In case someone wants to suggest it: A scatterplot won't do the job for me, I want to be able to see easily how many correlations are between 0.3 - 0.5 or -.3, -.5 etc.
I appreciate any help, as even google fails to help me with this problem!
I need to make a histogram with the Pearson correlations I calculated with SPSS to see how the correlations are distributed. When I click on the output table I am somewhat able to make one, but this includes also the correlation of Variable1 with Variable1 (=1) which I of course do not need.
Is there any other possibility? Can I make a new variable with the Pearson correlations for the datas? In case someone wants to suggest it: A scatterplot won't do the job for me, I want to be able to see easily how many correlations are between 0.3 - 0.5 or -.3, -.5 etc.
I appreciate any help, as even google fails to help me with this problem!
0
reply
Report
#2
(Original post by Papayaa)
Hello,
I need to make a histogram with the Pearson correlations I calculated with SPSS to see how the correlations are distributed. When I click on the output table I am somewhat able to make one, but this includes also the correlation of Variable1 with Variable1 (=1) which I of course do not need.
Is there any other possibility? Can I make a new variable with the Pearson correlations for the datas? In case someone wants to suggest it: A scatterplot won't do the job for me, I want to be able to see easily how many correlations are between 0.3 - 0.5 or -.3, -.5 etc.
I appreciate any help, as even google fails to help me with this problem!
Hello,
I need to make a histogram with the Pearson correlations I calculated with SPSS to see how the correlations are distributed. When I click on the output table I am somewhat able to make one, but this includes also the correlation of Variable1 with Variable1 (=1) which I of course do not need.
Is there any other possibility? Can I make a new variable with the Pearson correlations for the datas? In case someone wants to suggest it: A scatterplot won't do the job for me, I want to be able to see easily how many correlations are between 0.3 - 0.5 or -.3, -.5 etc.
I appreciate any help, as even google fails to help me with this problem!
You're right in saying you do not need the V1=V1 correlations - these should form a diagonal line across the table, and you can also disregard either everything above that or everything below that as it is repeated information.
It will probably be more meaningful to find and report which of the correlations are significant than how many of each strength.
0
reply
(Original post by Magdatrix >_<)
How many variables do you have? What hypotheses are you testing?
You're right in saying you do not need the V1=V1 correlations - these should form a diagonal line across the table, and you can also disregard either everything above that or everything below that as it is repeated information.
It will probably be more meaningful to find and report which of the correlations are significant than how many of each strength.
How many variables do you have? What hypotheses are you testing?
You're right in saying you do not need the V1=V1 correlations - these should form a diagonal line across the table, and you can also disregard either everything above that or everything below that as it is repeated information.
It will probably be more meaningful to find and report which of the correlations are significant than how many of each strength.
I have 4 variables, 2 for timeframe a and 2 for timeframe b, which I correlate with another variable for both times frames. I therefore do 4 correlations, split by 40 participants. I want to test the hypothesis if the majority of correlations will be moderately positive.
In my written report I do exactly what you suggested: mention the significant ones, how many, of which strength. I nevertheless wanted to make a graph (honestly just because my supervisor asked for this specific graph) where I highlight the significant ones.
Or would you suggest to omit the insignificant results completely? Because then I do not have many correlations left, with which I can either make a graph manually or for which a graph is of no value.
0
reply
X
Page 1 of 1
Go to first unread
Skip to page:
Quick Reply
Today on TSR
-
Ofqual consultationTeachers to use 'mini-exams' and other work to decide grades
- Is it too late to do well in my A-levels?
- What Harry Potter house are you in?
- 2021 wellbeing blog launch
- Student tips on lockdown mental health