Multivariate analysis of categorical data spss

[John F Hall]

Heather

 

Finding me on YouTube must have been very indirect as I have no entries there at all

 

I’m not a statistician and I only really deal with SPSS.  Excel is something I only use to copy/paste occasionally or read other people’s data into SPSS.  I can’t easily scroll across the full Excel file you sent as there are no slider bars at the side or bottom.

 

Do you have access to SPSS?  If so, you would be better addressing your query to the SPSS-X list [hidden email]

 

The experts there advise NEVER to use Excel for statistical analysis.  To get best advice from them you need to be very precise about what you want and give examples of data and/or syntax.  The list does NOT accept attachments or embedded links, so data and syntax examples need to be carefully specified as text or from SPSS output.

 

I suggest you check out:

https://www.google.fr/search?q=multivariate+analysis+of+categorical+data+spss&oq=multivariate+analysis+of+categorical+data+spss&aqs=chrome..69i57.1967j0j7&sourceid=chrome&espv=210&es_sm=93&ie=UTF-8

. . which lists several relevant on-line resources.

 

I have managed to extract the following table from your Excel file:

 

Correlations Original Variables
	Childhood SES	Native English Speaker	Age	Gender 2	Ethnicity Category	First Generation Status 2	Highest Degree Earned	Major Category	Employment by Type	Employment by Healthcare	Accepted Status
Childhood SES	1.000	-.362	-.218	.008	.234	.417	.112	-.015	-.058	.011	-.115
Native English Speaker	-.362	1.000	.094	-.001	-.358	-.158	-.099	.043	.019	-.009	.099
Age	-.218	.094	1.000	-.070	-.011	-.135	-.411	.098	.260	.076	.184
Gender 2	.008	-.001	-.070	1.000	.018	.017	.023	.010	-.067	-.013	-.035
Ethnicity Category	.234	-.358	-.011	.018	1.000	.134	.100	-.047	-.013	-.026	-.099
First Generation Status 2	.417	-.158	-.135	.017	.134	1.000	.078	-.018	-.021	-.026	-.077
Highest Degree Earned	.112	-.099	-.411	.023	.100	.078	1.000	-.159	-.144	-.029	-.102
Major Category	-.015	.043	.098	.010	-.047	-.018	-.159	1.000	.007	-.010	-.013
Employment by Type	-.058	.019	.260	-.067	-.013	-.021	-.144	.007	1.000	.349	.138
Employment by Healthcare	.011	-.009	.076	-.013	-.026	-.026	-.029	-.010	.349	1.000	.076
Accepted Status	-.115	.099	.184	-.035	-.099	-.077	-.102	-.013	.138	.076	1.000
Dimension	1	2	3	4	5	6	7	8	9	10	11
Eigenvalue	2.159	1.516	1.185	1.026	.993	.942	.843	.693	.610	.537	.495

 

. . but this doesn’t really tell me anything.  It looks as if you are using survey data of some sort: which survey is it?  Is the survey available as a saved file in SPSS format?  There’ll be a way of feeding the correlation matrix from this into SPSS, but it may need additional information.

 

You can access the CATREG facility from inside SPSS via:

 

Help >> Command Syntax Reference:

 

Further than that, I can’t be more help at this stage, but I’m copying this to list in case anyone needs a break from seasonal festivities.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  : :  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

From: [hidden email] [mailto:[hidden email]]
Sent: 30 December 2013 18:58
To: [hidden email]
Subject: multivariate analysis of categorical data spss

 

Greetings!

I found you on YouTube and I was wondering if you might be able to assist me.  I'm finishing up my PHD dissertation and I'm getting an error message with my output utilizing categorical multiple regression analysis. 

 

1. Specifically, I don't see my dependent variable on the output
2. I don't understand discretization
3. How do I fix the "too many interior knots?"

Any assistance you can provide would be GREATLY appreciated!

 

Sincerely,

Heather Petrelli

---

This email is free from viruses and malware because avast! Antivirus protection is active.

[David Marso]

John,
<SNARK>
Don't you just love it when total strangers send you unsolicited consulting requests and don't even offer to pay you anything?
</SNARK>
 I would suggest to OP to review the Algorithms documentation on whatever procedure you are running!!!  
D.

[John F Hall]

Not sure if my original reply to Heather was circulated, but it’s on Nabble.  She sent me an Excel xlsm file which had syntax similar to SPSS CATREG, from which I extracted a correlation matrix (see Nabble)

 

I replied (1 Jan 2014):

 

David Marso, Bruce Weaver and others far better qualified than me to answer your CATREG query.  Excel sheet no good to me, but I do have SPSS 21: can you send [off-list] your *.sav file?  Also, which survey are you using?

 

 

From: Heather [mailto:[hidden email]]
Sent: 01 January 2014 18:12
To: [hidden email]
Subject: Re: Multivariate analysis of categorical data spss

 

Hi John,

Thank you so much; I am honored that you took the time to write back.  I will follow up with the resources you provided.  Incidentally, these data were run through SPSS.  I just exported them in Excel because some people do not have immediate access to SPSS.  At any rate, this was survey research that also incorporated the Academic Motivation Scale.  My DV is the Self-Determination Index, a score that ranges from -18 to 18.  My independent variables are both categorical (gender, age, SES, major, etc) and continuous (GPA and standardized testing scores).  The output I sent to you is from the SPSS CATREG feature and I can't figure out how to tell the program that my DV is continuous and will it properly  show up in the results if I use the CATREG feature   I also can't find a way to do hierarchical and stepwise regressions using the CATREG feature. 

 

At any rate, thanks again, I will attempt to access the list-serve.  I've already identified many online resources, but so far, none have been able to answer my very specific questions.

 

Thanks again,

Heather