Help where are the equations?
Locked
Help where are the equations?
 
|
In earlier SPSS versions one Ould access equation used in each procedure
Now I can’t find them
In particular want Brown-Forsythe as SPSS does not give same df as in Wikipedia caution
best
Diana
=====================
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
____________
University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK +44 (0) 208 444 2081 Save our in-boxes! http://emailcharter.org
__________________ |
Re: Help where are the equations?
Administrator
|
Good morning Diana. Try clicking Help > Documentation in PDF format. Doing
that took me to this page for v25 (the version I have): https://www.ibm.com/support/pages/node/618179 If you use a Ctrl-F search on <algo>, you'll quickly find a PDF with the algorithms. Cheers, Bruce Kornbrot, Diana wrote > In earlier SPSS versions one Ould access equation used in each procedure > Now I can’t find them > In particular want Brown-Forsythe as SPSS does not give same df as in > Wikipedia caution > best > Diana > ____________ > University of Hertfordshire > College Lane, Hatfield, Hertfordshire AL10 9AB, UK > +44 (0) 208 444 2081 > +44 (0) 7403 18 16 12 > d.e.kornbrot@.ac > <mailto: > d.e.kornbrot@.ac > > > http://dianakornbrot.wordpress.com/ > skype: kornbrotme > Save our in-boxes! http://emailcharter.org > __________________ > > > > > > > > > ===================== > To manage your subscription to SPSSX-L, send a message to > LISTSERV@.UGA > (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 ----- -- 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. -- Sent from: http://spssx-discussion.1045642.n5.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
--
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/). |
Re: Help where are the equations?
|
Administrator
|
In reply to this post by Kornbrot, Diana
Diana, as I thought about your post some more, it occurred to me that there
could be confusion because the term Brown-Forsythe test is used to describe two completely different tests. One is a test for homogeneity of variance, similar to Levene's test. The other is a robust F-test for equality of means, similar to Welch's F. The only Wikipedia page I can find is about the test of variance homogeneity. But the Brown-Forsythe test available via ONEWAY is a robust F-test for equality of means. Perhaps this discrepancy is at the heart of your problem. https://en.wikipedia.org/wiki/Brown%E2%80%93Forsythe_test Here is an old ResearchGate discussion I participated in that brought some of these issues to light. Another poster who used SAS kept talking about the Brown-Forsythe test as a test of homogeneity of variance, and that confused me at first, because I thought of it as the robust F-test from ONEWAY. https://www.researchgate.net/post/what_is_the_assumption_for_Brown_Forsythe_test_and_Welch_test I've attached here the same syntax file I attached to my post to RG on 21-Feb-2018. It includes code to calculate some of the tests of homogeneity of variance "by hand" to clarify exactly how they work. (The uploaded Excel file shows the output for readers who do not have access to SPSS.) HTH. Levene_&_Brown-Forsythe_tests.sps <http://spssx-discussion.1045642.n5.nabble.com/file/t7186/Levene_%26_Brown-Forsythe_tests.sps> Levene-Brown-Forsythe-examples.xls <http://spssx-discussion.1045642.n5.nabble.com/file/t7186/Levene-Brown-Forsythe-examples.xls> Kornbrot, Diana wrote > In earlier SPSS versions one Ould access equation used in each procedure > Now I can’t find them > In particular want Brown-Forsythe as SPSS does not give same df as in > Wikipedia caution > best > Diana > ____________ > University of Hertfordshire > College Lane, Hatfield, Hertfordshire AL10 9AB, UK > +44 (0) 208 444 2081 > +44 (0) 7403 18 16 12 > d.e.kornbrot@.ac > <mailto: > d.e.kornbrot@.ac > > > http://dianakornbrot.wordpress.com/ > skype: kornbrotme > Save our in-boxes! http://emailcharter.org > __________________ > > > > > > > > > ===================== > To manage your subscription to SPSSX-L, send a message to > LISTSERV@.UGA > (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 ----- -- 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. -- Sent from: http://spssx-discussion.1045642.n5.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
--
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/). |
Re: Help where are the equations?
|
Administrator
|
Here is the original article describing the Brown-Forsythe test for equality
of means (the one available via ONEWAY): Brown, M., & Forsythe, A. (1974). The Small Sample Behavior of Some Statistics Which Test the Equality of Several Means. Technometrics, 16(1), 129-132. doi:10.2307/1267501 JSTOR: https://www.jstor.org/stable/1267501?seq=1#metadata_info_tab_contents And here is the article describing the test of homogeneity of variance: Brown, M., & Forsythe, A. (1974). Robust Tests for the Equality of Variances. Journal of the American Statistical Association, 69(346), 364-367. doi:10.2307/2285659 JSTOR: https://www.jstor.org/stable/2285659?seq=1#metadata_info_tab_contents Search results (and my previous experience) suggest that people frequently say "Brown-Forsythe" test without specifying which test they mean, and that causes no end of confusion--especially when people using different stats packages try to have a conversation. Bruce Weaver wrote > Diana, as I thought about your post some more, it occurred to me that > there > could be confusion because the term Brown-Forsythe test is used to > describe > two completely different tests. One is a test for homogeneity of > variance, > similar to Levene's test. The other is a robust F-test for equality of > means, similar to Welch's F. The only Wikipedia page I can find is about > the test of variance homogeneity. But the Brown-Forsythe test available > via > ONEWAY is a robust F-test for equality of means. Perhaps this discrepancy > is at the heart of your problem. > > https://en.wikipedia.org/wiki/Brown%E2%80%93Forsythe_test > > Here is an old ResearchGate discussion I participated in that brought some > of these issues to light. Another poster who used SAS kept talking about > the Brown-Forsythe test as a test of homogeneity of variance, and that > confused me at first, because I thought of it as the robust F-test from > ONEWAY. > > https://www.researchgate.net/post/what_is_the_assumption_for_Brown_Forsythe_test_and_Welch_test > > I've attached here the same syntax file I attached to my post to RG on > 21-Feb-2018. It includes code to calculate some of the tests of > homogeneity > of variance "by hand" to clarify exactly how they work. (The uploaded > Excel > file shows the output for readers who do not have access to SPSS.) > > HTH. ----- -- 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. -- Sent from: http://spssx-discussion.1045642.n5.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
--
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/). |
Re: Help where are the equations?
|
|
In reply to this post by Kornbrot, Diana
If you are looking for Algorithms that used to be accessible from SPSS help, you can check the following best wishes Forcheh On Thu, Jul 2, 2020 at 5:09 AM Kornbrot, Diana <[hidden email]> wrote:
|
Re: Help where are the equations?
|
|
In reply to this post by Bruce Weaver
Hi Bruce, I followed the links and find very interesting and informative discussions - especially your contributions - about the different tests and assumptions. Such discussions should be shared with students taking statistical inference courses (if they don't increase confusion and failure rates hahaha). Thanks for sharing. Forcheh On Thu, Jul 2, 2020 at 4:48 PM Bruce Weaver <[hidden email]> wrote: Diana, as I thought about your post some more, it occurred to me that there |
Re: Help where are the equations?
|
|
In reply to this post by Nkem Ntonghanwah
Many thanks
Apparently it is for v24 although it may include amendments for later versions
Who knows?
IBM is a DISGRACE in not making algorithms immediately findable
AS JASP improves am likely to recommend my Uni gradually withdraws from SPSS
____________
University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK +44 (0) 208 444 2081 Save our in-boxes! http://emailcharter.org
__________________ |
Re: Help where are the equations?
|
Administrator
|
This post was updated on .
For some period of time, the med school I'm connected to was not making any
statistical software available to faculty, staff or students. During that period, to meet an accreditation standard, we had to add a lab in which students did some simple statistical analyses on data they collected. Because there was no software available to them via the School, I had them do it using JASP. To be fair, JASP has some nice features, and produces some nice graphical output, etc. But in its current state, I would never recommend it as a tool for serious data analysis. JASP has (at least) two glaring deficiencies, I think: 1. There is no (easy) way to generate code. And for serious statistical analysis, code is essential for reproducibility and for documentation of exactly what was done. 2. Its data management capabilities are not nearly as extensive as they ought to be for serious data analysis, IMO. So if you're going to recommend that your Uni move away from SPSS, I don't think JASP is the answer. Having said all that, I do agree that it would be very helpful if the Help > Documentation in PDF format option in the menu gave some indication that this is where to find documentation about the algorithms. Bruce Kornbrot, Diana wrote > Many thanks > > or https://www.ibm.com/support/pages/ibm-spss-statistics-25-documentation > > This one worked > Apparently it is for v24 although it may include amendments for later > versions > Who knows? > IBM is a DISGRACE in not making algorithms immediately findable > AS JASP improves am likely to recommend my Uni gradually withdraws from > SPSS > > --- snip --- ----- -- Bruce Weaver bweaver@lakeheadu.ca 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. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@LISTSERV.UGA.EDU (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
--
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 Kornbrot, Diana
If you go to Help > Topics and search for algorithms, you will find IBM SPSS Statistics V 27 Documentation and on that page there is a direct link to the Algorithms Guide, which is almost 1400 pages of documentation. On Fri, Jul 3, 2020 at 4:39 AM Kornbrot, Diana <[hidden email]> wrote:
|
Re: Help where are the equations?
|
|
In reply to this post by Kornbrot, Diana
Diana wrote "AS JASP improves am likely to recommend my Uni gradually withdraws from SPSS" Hi Diana, I am not connected with SPSS, but I am sure if users demanded that algorithms be included among documents - such as Command Syntax in help unido, It shouldn't be difficult for them to do so. Of course, some changes are not reversible - as what happens when they changed the custom tables menu after version 13. I wrote to SPSS and complained that some important features that were available in the old custom tables - especially what used to be called General tables were missing in the new customs tables, and they disagreed with me, but failed to show me how I could use the new format to regenerate my old tables -- one suggestion was for me to use Python which I found unsatisfying. However, I continue to find many good things with SPSS. Forcheh On Fri, Jul 3, 2020 at 6:39 AM Kornbrot, Diana <[hidden email]> wrote:
|
«
Return to SPSSX Discussion
|
1 view|%1 views
| Free forum by Nabble | Edit this page |