NPAR TESTS Question
Locked
 
|
Colleagues:
I ran a Friedman test using NPAR TESTS. . My syntax was as follows: NPAR TESTS /FRIEDMAN=V83 V102 V121 V140 /FRIEDMAN=V84 V103 V122 V141 /FRIEDMAN=V85 V104 V123 V142. In my output, I get a box that gives me the results. like this:
My question relates to the calculation of N in the Test Statistics box. What is it based on? My data set has between 22 and 24 subjects contributing to each of the three variables in the analysis. Why do I get the N=2 and N=3 in the stats box? Is this related, perhaps, to how I am handling missing data (the default)? Now the really important question: How could I have found the answer to this fairly simple question in the SPSS help documentation or elsewhere? I miss the old SPSS Manual! David David W. Thompson, PhD, ABPP Diplomate in Forensic Psychology American Board of Professional Psychology Deputy Director Walworth Co. Dept. of Health and Human Services 262-741-3232 (voice) 262-741-3217 (fax) NOTICE OF CONFIDENTIALITY: This e-mail and any files transmitted with it may contain information that is privileged, confidential, and exempt from disclosure under applicable laws. This communication is intended for the sole use of the individual or entity to which it is addressed. Dissemination, forwarding, printing, or copying of this e-mail and any files transmitted with it without the consent of the sender is strictly prohibited. If you have received this e-mail in error, please do not distribute it. Please notify the sender by e-mail at the address shown and delete the original message. Thank you. |
||||||||||||||||||||||||||||||||||||
|
|
A couple of points:
(1) Below the output you say:
>My data set has between 22 and 24 subjects
contributing
>to each of the three variables in the
analysis
But you have four (4) variables. Did you
mis-speak or do
you have an extra variable in the analysis (which
might be
throwing it off).
(2) Not clear to me where you would find the answer
to your
question in the SPSS documentation but I would
suggest
having, say, Andy Field's "Discovering Statistics
Using IBM
SPSS Statistics" as a possible guide. In the
4th edition he
covers the Friedman test on pages
251-256.
-Mike Palij
New York University
|
||||||||||||||||||||||||||||||||||||
Administrator
|
In reply to this post by David Thompson
How much missing data have you got (or not got, I suppose)? What are the valid listwise Ns for the following DESCRIPTIVES commands?
DESCRIPTIVES V83 V102 V121 V140. DESCRIPTIVES V84 V103 V122 V141. DESCRIPTIVES V85 V104 V123 V142.
--
Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." PLEASE NOTE THE FOLLOWING: 1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. 2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/). |
|
|
In reply to this post by David Thompson
Hi David, The Freidman is a complete-case method of analysis, so each test will only include cases that have valid data for all four variables. You’re right that this
doesn’t appear to be explicitly mentioned in the documentation that I could see, but with most stats methods it’s probably safest to assume a method will use complete cases only unless it is explicitly described otherwise. Hope this helps. Kylie. From: SPSSX(r) Discussion [mailto:[hidden email]]
On Behalf Of David Thompson Colleagues:
|
||||||||||||||||||||||||||||||||||||
Administrator
|
In reply to this post by Bruce Weaver
@OP
"Not got!" LOL PMP! Way back in the ancient times we didn't even have the data editor. Maybe take a peek? and DESC? <RANT> Why don't people bother to look at their data prior to processing it? </RANT> ------------------------
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me. --- "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis." Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?" |
Automatic reply: NPAR TESTS Question
|
|
I will be out of my office until March 12, 2013.
|
|
|
In reply to this post by Bruce Weaver
Thank you to those of you who responded
to my inquiry. I found most of the comments helpful and informative.
Yes, I mis-spoke when I said that I had three variables going into the analysis, and then listed four! The primary investigator asked me to throw an additional one in at the last moment. I did not know that the program would consider only cases with valid data across all four variables. I ran the DESCRIPTIVES, and see that the valid N is indeed 2 or 3, depending on the variables, even though my data set is larger. Had I known that it would exclude cases listwise, I would have checked it earlier. I miss the old SPSS manual, where I would have learned more about the procedure before selecting it. I will have to revisit my approach to analyzing the data. David David W. Thompson, PhD, ABPP Diplomate in Forensic Psychology American Board of Professional Psychology Deputy Director Walworth Co. Dept. of Health and Human Services 262-741-3232 (voice) 262-741-3217 (fax) NOTICE OF CONFIDENTIALITY: This e-mail and any files transmitted with it may contain information that is privileged, confidential, and exempt from disclosure under applicable laws. This communication is intended for the sole use of the individual or entity to which it is addressed. Dissemination, forwarding, printing, or copying of this e-mail and any files transmitted with it without the consent of the sender is strictly prohibited. If you have received this e-mail in error, please do not distribute it. Please notify the sender by e-mail at the address shown and delete the original message. Thank you. From: Bruce Weaver <[hidden email]> To: [hidden email] Date: 03/05/2013 09:36 PM Subject: Re: NPAR TESTS Question Sent by: "SPSSX(r) Discussion" <[hidden email]> How much missing data have you got (or /not/ got, I suppose)? What are the valid listwise Ns for the following DESCRIPTIVES commands? DESCRIPTIVES V83 V102 V121 V140. DESCRIPTIVES V84 V103 V122 V141. DESCRIPTIVES V85 V104 V123 V142. David Thompson wrote > Colleagues: > > I ran a Friedman test using NPAR TESTS. > . > My syntax was as follows: > > NPAR TESTS > /FRIEDMAN=V83 V102 V121 V140 > /FRIEDMAN=V84 V103 V122 V141 > /FRIEDMAN=V85 V104 V123 V142. > > In my output, I get a box that gives me the results. like this: > > Test Statisticsa > N > 2 > Chi-Square > 5.842 > df > 3 > Asymp. Sig. > .120 > a. Friedman Test > > > Test Statisticsa > N > 3 > Chi-Square > 1.200 > df > 3 > Asymp. Sig. > .753 > a. Friedman Test > > > Test Statisticsa > N > 3 > Chi-Square > 5.880 > df > 3 > Asymp. Sig. > .118 > a. Friedman Test > > > My question relates to the calculation of N in the Test Statistics box. > What is it based on? My data set has between 22 and 24 subjects > contributing to each of the three variables in the analysis. Why do I get > the N=2 and N=3 in the stats box? Is this related, perhaps, to how I am > handling missing data (the default)? > > Now the really important question: How could I have found the answer to > this fairly simple question in the SPSS help documentation or elsewhere? I > miss the old SPSS Manual! > > David > > David W. Thompson, PhD, ABPP > Diplomate in Forensic Psychology > American Board of Professional Psychology > > Deputy Director > Walworth Co. Dept. of Health and Human Services > 262-741-3232 (voice) 262-741-3217 (fax) > > NOTICE OF CONFIDENTIALITY: This e-mail and any files transmitted with it > may contain information that is privileged, confidential, and exempt from > disclosure under applicable laws. This communication is intended for the > sole use of the individual or entity to which it is addressed. > Dissemination, forwarding, printing, or copying of this e-mail and any > files transmitted with it without the consent of the sender is strictly > prohibited. If you have received this e-mail in error, please do not > distribute it. Please notify the sender by e-mail at the address shown and > delete the original message. Thank you. ----- -- Bruce Weaver [hidden email] http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." NOTE: My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. -- View this message in context: http://spssx-discussion.1045642.n5.nabble.com/NPAR-TESTS-Question-tp5718394p5718395.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 |
|
|
The Command Syntax Reference is a PDF (syntaxreference.pdf) available from the Help Menu. It contains the syntax reference and covers the missing value handling for all procedures. It is stored in the root directory where SPSS is installed. From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of David Thompson Thank you to those of you who responded to my inquiry. I found most of the comments helpful and informative.
|
|
|
In reply to this post by David Thompson
David, There are alternative statistical procedures offered in SPSS which are capable of handling data which are missing at random (e.g., MIXED, GENLIN, GENLINMIXED) without having to discard all data from a subject. But I would need to know more about your research question, design, variables, etc. to offer up code.
The Friedman test, from what I recall, is intended to test for significant differences between three or more related groups on ordinal level data. I'm rarely fond of these types of tests. If you are interested in further consultation via SPSS-L, could you tell us why you are not fitting a repeated measures ANOVA, along with the other questions I just posed? Best, Ryan On Thu, Mar 7, 2013 at 10:41 AM, David Thompson <[hidden email]> wrote: Thank you to those of you who responded to my inquiry. I found most of the comments helpful and informative. |
|
|
In reply to this post by Vi Beadle
Also note that the documentation for v21
is at http://pic.dhe.ibm.com/infocenter/spssstat/v21r0m0/index.jsp
in HTML format. The syntax topic for NPAR TESTS, specifically, is
at: http://pic.dhe.ibm.com/infocenter/spssstat/v21r0m0/topic/com.ibm.spss.statistics.help/syn_npar_tests.htm.
I don't think the NPAR TESTS documentation specifically mentions
that records are deleted casewise because that's the typical handling of
missing values throughout the product.
All the PDFs are at: http://www-01.ibm.com/support/docview.wss?uid=swg27024972 Alex |
| Free forum by Nabble | Edit this page |