Making SPSS better
Locked
 
|
Will be back in US in a couple weeks and respond in more detail. Meanwhile, it would be great to have the same tests in different procedure have similar "looks".
see the archives for my name and "pretty" for syntax related suggestions I have made over the last few decades. to help reinforce good habits: When a new variable in encountered in syntax here should be an option for "variable not yet defined" that brings up box for print formats, valid value labels, missing values, missing value labels, and variable label (perhaps with the formula as default var label). an option to set exit from GUI that asks :Are you sure you do not want to PASTE, when OK is keyed. ---- An option from GUI boxes to PASTE & GO. In my opinion Human factors are on of SPSS's strengths and they should work on those more.
Art Kendall
Social Research Consultants |
|
|
My suggestions are based on seeing people "shoot themselves in the foot" with many packages since 1971.
Art Kendall
Social Research Consultants |
|
In reply to this post by Bruce Weaver
Bruce The point I was making is that certain kinds of students and researchers need to start with something like elaboration as an aid to understanding the logic, and perhaps even using, more advanced methods. I have come across too many people, especially in psychology, who dive straight into means and correlation matrices without ever looking at frequency distributions and contingency tables. Path diagrams are a good starting point for both elaboration and statistical modelling. Incidentally, I Jim Davis (author of Elementary Survey Analysis https://www.amazon.com/Elementary-Analysis-Prentice-Hall-methods-science/dp/0132605473 ) once referred (at a summer school on survey methods) to work at Johns Hopkins by William Huggins (electrical engineer) and Doris Entwisle (sociologist) using electrical circuit boards as examples of path models (with voltmeters, resistors and rheostats to vary inputs and currents along different wires/paths). http://pages.jh.edu/~gazette/2001/aug2001/20huggin.html ) There is no bibliography in the Davis book: does anyone know of a specific reference or if such a circuit board was actually built? 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 Cc: ASA Survey Research Methods Section -----Original Message----- John, I understand that you are trying to make a point about the usefulness of basic descriptive statistics, cross-tabulations and graphs. No one disagrees with that, and frequent reminders of their usefulness are probably helpful. But the usefulness of those things does not negate the usefulness of more advanced methods. We need both. Re David's thesis, notice that he indicated neither the discipline nor the level (undergrad, Masters, PhD) of the thesis. I /hope/ that you are not seriously suggesting that every thesis, at every level and in every discipline should be written in such a way that it can be /easily/ understood by undergraduates in Sociology, Political Science, Social Administration etc. That's just silly. John F Hall wrote > Yes, but did anyone read it? If so, how many of them were > undergraduates in Sociology, Political Science, Social Administration > etc. and how many of them understood it? As my old boss, the late Dr > Mark Abrams (1906 - > 1994) http://surveyresearch.weebly.com/mark-abrams.html (one of the > founding fathers of market and survey research) once said, "If it's worth > saying, you can say it with percentages!" Today he might have conceded > the same for charts in SPSS. > > John F Hall (Mr) > [Retired academic survey researcher] > > Email: > johnfhall@ > > Website: www.surveyresearch.weebly.com SPSS start page: > www.surveyresearch.weebly.com/1-survey-analysis-workshop > > > -----Original Message----- > From: SPSSX(r) Discussion [mailto: > ] On Behalf Of David Marso > Sent: 07 July 2016 12:54 > To: > Subject: Re: Making SPSS better > > I suspect most people with minimal common sense and training would > utilize LOG LINEAR models to parse this sort of business. Not so fancy. > Civilized people such as myself have been using them for over 30 > years. I used them to great effect for my thesis back in 1986. ----- -- Bruce Weaver 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/Making-SPSS-better-tp5732642p5732665.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 |
Administrator
|
In reply to this post by Christopher Stride
Related to that point, a /PLOT sub-command would be a nice addition to GENLIN, MIXED and maybe even GENLINMIXED (cf. UNIANOVA).
--
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 spss.giesel@yahoo.de
I have collected a short list wishings for coming versions. I do believe it is important to make the syntax output nice/readable to help non-programmers to understand what is actually going on behinde the GUI.
Some of these may have been implemented, as I have not updated it for a while. * Option to get Leading zeroes in range -1...1 : SET ZEROLEAD ON|OFF * Selectable ISO 8601 sortable date SDATE10 format 'yyyy-mm-dd' or even 'yyyymmdd' (with/without dashes), maybe even with 2-digit year 'yy-mm-dd' or 'yymmdd' within EPOCH * Date constants; Compute dSurvey=2012/12/31 (or 20121231, 2012-12-31). * Variable manipulation tools in syntax; Var label, Value label; min, max * SPSS Standard save location NOT in (Vista/Win7) protected 'Program Files'-folder, but in UserProfile's folder. (templates, looks, etc). * Lockable toolbars; shouldn't change position after a near miss. * Enhanced label control in graphs * Enhanced legend control in graphs * Graphs with log axis: minor tickmarks * Chart Editor: when closing window: 'Accept/Cancel changes'. Maybe even undo levels... * Chart Editor: axis number format Scaling Factor should be multiplying, not dividing. * Multiple predefined/cusom defined Color palettes for charts (Color/Gray) * Syntax coloring within macro definitions, and maybe a different background color * Debugged syntax coloring... * Syntax help to insert '=' when demanded after keyword * Syntax help to insert '.' when appropriate ('exe' --> 'EXECUTE.') * Syntax Command completion even when multiple words (DATASET..., END...) * A modern Syntax Editor, compare UltraEdit or NotepadPlus and others with macros/search/replace: lots of bells & whistles (OR allowing integration of an external editor!); - Auto indent - column mode editing - too easy to accidentaly move selected text around, - terrible undo-function (splits action in several steps) * Loading a syntax statement into a dialog box for easier changing/adjusting (yes I know; frontend/backend). Maybe a flag raised if a non-dialog option was encountered. * The TEXT tool integrated in syntax with correct PAGEBREAK placement, multiline comments and /COMMENT to place text-item immediately after title-item in previous command. Also allowing Heading without comment text. Especially helpful when generating lots of reports on a regular basis !* MODIFY OUTPUT tool integrated in syntax and easier use. Hide entire command entry * Bugfix: correct names in command id's, subtables and labels! (e.g. ALTER TYPE) !* MODIFY TABLES tool integrated in syntax and easier use * CROSSTAB option /refcat(1) to get odds ratios intuitively correct when (0|1) coding (cf LOGREG) * VARIABLE SET definable in syntax, to recreate SET when new dataset is loaded from other source. * AGGREGATE : paste generates incorrect OVERWRITEVARS (should be OVERWRITE) * Consistent grammar through syntax commands |
|
SDATE format has been available for some time. SDATE input values without delimiters and leading zeroes for display of decimal values were both added in release 24. data list list /datevar (sdate10). begin data 1986-10-28 19861028 1986/10/28 end data. set leadzero yes. compute x=.1. list. Sent from Mail for Windows 10 From: [hidden email] I have collected a short list wishings for coming versions. I do believe it is important to make the syntax output nice/readable to help non-programmers to understand what is actually going on behinde the GUI. Some of these may have been implemented, as I have not updated it for a while. * Option to get Leading zeroes in range -1...1 : SET ZEROLEAD ON|OFF * Selectable ISO 8601 sortable date SDATE10 format 'yyyy-mm-dd' or even 'yyyymmdd' (with/without dashes), maybe even with 2-digit year 'yy-mm-dd' or 'yymmdd' within EPOCH * Date constants; Compute dSurvey=2012/12/31 (or 20121231, 2012-12-31). * Variable manipulation tools in syntax; Var label, Value label; min, max * SPSS Standard save location NOT in (Vista/Win7) protected 'Program Files'-folder, but in UserProfile's folder. (templates, looks, etc). * Lockable toolbars; shouldn't change position after a near miss. * Enhanced label control in graphs * Enhanced legend control in graphs * Graphs with log axis: minor tickmarks * Chart Editor: when closing window: 'Accept/Cancel changes'. Maybe even undo levels... * Chart Editor: axis number format Scaling Factor should be multiplying, not dividing. * Multiple predefined/cusom defined Color palettes for charts (Color/Gray)
* Syntax coloring within macro definitions, and maybe a different background color * Debugged syntax coloring... * Syntax help to insert '=' when demanded after keyword * Syntax help to insert '.' when appropriate ('exe' --> 'EXECUTE.') * Syntax Command completion even when multiple words (DATASET..., END...) * A modern Syntax Editor, compare UltraEdit or NotepadPlus and others with macros/search/replace: lots of bells & whistles (OR allowing integration of an external editor!); - Auto indent - column mode editing - too easy to accidentaly move selected text around, - terrible undo-function (splits action in several steps) * Loading a syntax statement into a dialog box for easier changing/adjusting (yes I know; frontend/backend). Maybe a flag raised if a non-dialog option was encountered. * The TEXT tool integrated in syntax with correct PAGEBREAK placement, multiline comments and /COMMENT to place text-item immediately after title-item in previous command. Also allowing Heading without comment text. Especially helpful when generating lots of reports on a regular basis !* MODIFY OUTPUT tool integrated in syntax and easier use. Hide entire command entry * Bugfix: correct names in command id's, subtables and labels! (e.g. ALTER TYPE) !* MODIFY TABLES tool integrated in syntax and easier use
* CROSSTAB option /refcat(1) to get odds ratios intuitively correct when (0|1) coding (cf LOGREG)
* VARIABLE SET definable in syntax, to recreate SET when new dataset is loaded from other source. * AGGREGATE : paste generates incorrect OVERWRITEVARS (should be OVERWRITE) * Consistent grammar through syntax commands -- View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Making-SPSS-better-tp5732642p5732674.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 |
|
|
Also iOS 8601 date format, chart looks for color schemes are available. Log scales have been available for many years.
Not sure what is meant about MODIFY OUTPUT, MODIFY TABLES, and TEXT, but the OUTPUT MODIFY native command, modeled on the first two of these with a nice gui was added in V22. On Friday, July 8, 2016, Rick Oliver <[hidden email]> wrote:
-- ===================== 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 |
Re: Making SPSS better, either, or +BOTH
|
|
Small interface issue, but save much time
Frequently SPSS offers either, or but not BOTH - when user does want BOTH. So have to run twice which is a pain
eg1 EXPORT
EXPORT data to EXCEL for nominal.
Options: number code OR label. I want BOTH
eg2 Generalised linear or MIXED logit output MEANS
Options, original metric or logit. I want BOTH
NB 23 seems to have lost option to specify number format in pivot tables when double clicking - yuk
INTERFACE as well as calculations matters, especially as this may ow be main reason fro spss, costly, rather than R free
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
________________________________________
Professor Diana
Kornbrot
Work
University of
Hertfordshire
College Lane,
Hatfield, Hertfordshire AL10 9AB, UK
+44 (0) 170 728 4626
skype:
kornbrotme
Home
19 Elmhurst Avenue
London N2 0LT,
UK
+44
(0) 208 444 2081
|
|
|
In reply to this post by PRogman
on crosstabs surely would like confidence levels on row or column proportions. one has to also run generalised linear to get these.
are procedures need effect sizes, with possible option to CHOOSE effect size diana > On 9 Jul 2016, at 01:09, PRogman <[hidden email]> wrote: > > I have collected a short list wishings for coming versions. I do believe it > is important to make the syntax output nice/readable to help non-programmers > to understand what is actually going on behinde the GUI. > Some of these may have been implemented, as I have not updated it for a > while. > > * Option to get Leading zeroes in range -1...1 : SET ZEROLEAD ON|OFF > * Selectable ISO 8601 sortable date SDATE10 format 'yyyy-mm-dd' or even > 'yyyymmdd' (with/without dashes), maybe even with 2-digit year 'yy-mm-dd' or > 'yymmdd' within EPOCH > * Date constants; Compute dSurvey=2012/12/31 (or 20121231, 2012-12-31). > * Variable manipulation tools in syntax; Var label, Value label; min, max > * SPSS Standard save location NOT in (Vista/Win7) protected 'Program > Files'-folder, but in UserProfile's folder. (templates, looks, etc). > * Lockable toolbars; shouldn't change position after a near miss. > > * Enhanced label control in graphs > * Enhanced legend control in graphs > * Graphs with log axis: minor tickmarks > * Chart Editor: when closing window: 'Accept/Cancel changes'. Maybe even > undo levels... > * Chart Editor: axis number format Scaling Factor should be multiplying, > not dividing. > * Multiple predefined/cusom defined Color palettes for charts (Color/Gray) > > * Syntax coloring within macro definitions, and maybe a different > background color > * Debugged syntax coloring... > * Syntax help to insert '=' when demanded after keyword > * Syntax help to insert '.' when appropriate ('exe' --> 'EXECUTE.') > * Syntax Command completion even when multiple words (DATASET..., END...) > * A modern Syntax Editor, compare UltraEdit or NotepadPlus and others with > macros/search/replace: lots of bells & whistles (OR allowing integration of > an external editor!); > - Auto indent > - column mode editing > - too easy to accidentaly move selected text around, > - terrible undo-function (splits action in several steps) > * Loading a syntax statement into a dialog box for easier > changing/adjusting (yes I know; frontend/backend). > Maybe a flag raised if a non-dialog option was encountered. > > * The TEXT tool integrated in syntax with correct PAGEBREAK placement, > multiline comments > and /COMMENT to place text-item immediately after title-item in previous > command. Also > allowing Heading without comment text. Especially helpful when generating > lots of reports on a regular basis > > !* MODIFY OUTPUT tool integrated in syntax and easier use. Hide entire > command entry > * Bugfix: correct names in command id's, subtables and labels! (e.g. > ALTER TYPE) > > !* MODIFY TABLES tool integrated in syntax and easier use > > * CROSSTAB option /refcat(1) to get odds ratios intuitively correct when > (0|1) coding (cf LOGREG) > > * VARIABLE SET definable in syntax, to recreate SET when new dataset is > loaded from other source. > > * AGGREGATE : paste generates incorrect OVERWRITEVARS (should be OVERWRITE) > > * Consistent grammar through syntax commands > > > > > > -- > View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Making-SPSS-better-tp5732642p5732674.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 ________________________________________ Professor Diana Kornbrot Work University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK +44 (0) 170 728 4626 [hidden email] http://dianakornbrot.wordpress.com/ http://go.herts.ac.uk/Diana_Kornbrot skype: kornbrotme Home 19 Elmhurst Avenue London N2 0LT, UK +44 (0) 208 444 2081 ===================== 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 |
|
|
In reply to this post by spss.giesel@yahoo.de
- The techie's perspective -
Thanks for this discussion; from my point of view (born in 1986, working for a software manufacturer and being the older generation of the company), SPSS is a perfect example of a software product caught between "offering a standard toolset" (i.e. what has been there since decades) and "offering flexible functionality and integration" (i.e. innovation & automation). As customer expectations diverge and become more sophisticated, this conflict becomes even more obvious. * Syntax: SPSS syntax is "special", and you can only use it in SPSS (no portability whatsoever). By the time I try to convince people in our company to get their hands on SPSS syntax, they have thrown away the manual and start coding Python or R. So SPSS syntax is for SPSS only, and that is the key problem. Today, would you recommend your students to learn SPSS syntax? Why? (I am not saying it is useless, in fact it is the heart of SPSS and should also be in the future, but it does not meet any modern standards of accessibility and portability, due to its legacy nature. That is OK but not good.) * Integration and automation: IBM has done many steps in the right direction here; just imagine the possibilities of the OMS! It is easy to integrate SPSS in existing systems like report generators. By now, you can use R and Python. This is 100% great and covers all things necessary for a modern workflow (people need to be enabled, but this is 2016 and not the 1970ies). I used to play around with R, node.js and SPSS on a Windows VM. Worked great. What SPSS could improve is the internationalization aspect (better and more careful control of charsets and display options for date and number fields; e.g. SPSS should recommended its users to use unicode instead of any regional charset) * Performance: Yes, SPSS is indeed kind of slow. But most people can accept this as they really want the "standard toolset" of SPSS. The results are worth the waiting, at least for most use cases. :) Kind regards, Nils Glueck ===================== 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 |
|
|
SPSS is still specified in job adverts for social researchers, at least in the UK. Syntax is what gives SPSS its appeal to social researchers, since it is English language like and intuitive. I would always recommend syntax over the GUI when learning basic data management and analysis, particularly from questionnaire surveys.
Many academic institutions have moved over to Stata because of cost, but its syntax is no match for SPSS, nor is its output. R and Python are far too complex for social science beginners: they can move on to those later, once they have mastered SPSS. Except in rare cases, they will never become techies anyway, but the logic will be invaluable in their later careers, whether as researchers, managers or policy makers. 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 PS When can we have an additional option /cells in FREQUENCIES? We could then choose one or more of cou val per cum -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Nils Glueck Sent: 09 July 2016 15:32 To: [hidden email] Subject: Re: Making SPSS better - The techie's perspective - Thanks for this discussion; from my point of view (born in 1986, working for a software manufacturer and being the older generation of the company), SPSS is a perfect example of a software product caught between "offering a standard toolset" (i.e. what has been there since decades) and "offering flexible functionality and integration" (i.e. innovation & automation). As customer expectations diverge and become more sophisticated, this conflict becomes even more obvious. * Syntax: SPSS syntax is "special", and you can only use it in SPSS (no portability whatsoever). By the time I try to convince people in our company to get their hands on SPSS syntax, they have thrown away the manual and start coding Python or R. So SPSS syntax is for SPSS only, and that is the key problem. Today, would you recommend your students to learn SPSS syntax? Why? (I am not saying it is useless, in fact it is the heart of SPSS and should also be in the future, but it does not meet any modern standards of accessibility and portability, due to its legacy nature. That is OK but not good.) * Integration and automation: IBM has done many steps in the right direction here; just imagine the possibilities of the OMS! It is easy to integrate SPSS in existing systems like report generators. By now, you can use R and Python. This is 100% great and covers all things necessary for a modern workflow (people need to be enabled, but this is 2016 and not the 1970ies). I used to play around with R, node.js and SPSS on a Windows VM. Worked great. What SPSS could improve is the internationalization aspect (better and more careful control of charsets and display options for date and number fields; e.g. SPSS should recommended its users to use unicode instead of any regional charset) * Performance: Yes, SPSS is indeed kind of slow. But most people can accept this as they really want the "standard toolset" of SPSS. The results are worth the waiting, at least for most use cases. :) Kind regards, Nils Glueck ===================== 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 |
|
Administrator
|
John, I suspect that most Stata users would have that comment about Stata vs SPSS syntax exactly the other way around. ;-)
Unfortunately, such comparisons are frequently made by commentators who are very well-versed in one package but know very little (if anything) about the other package. Personally, I don't put much stock in such commentaries. I'm far more interested in the thoughts of people who are well-versed in both packages. Take Richard Williams, for example. He is a sociology prof at Notre Dame University (https://www3.nd.edu/~rwilliam/) who is quite well-versed in both SPSS and Stata. Many of his class notes show how to perform the relevant analyses using both packages. Williams makes the following comments in his notes on using Stata for OLS regression (https://www3.nd.edu/~rwilliam/stats1/OLS-Stata9.pdf), emphasis added: --- start of excerpt --- Rather than specify all options at once, like you do in SPSS, in Stata you often give a series of commands. In some ways this is more tedious, but it also gives you flexibility in that you don’t have to rerun the entire analysis if you think of something else you want. As the Stata 9 User’s Guide says (p. 43) “The user-interface model is type a little, get a little, etc. so that the user is always in control.” For the most part, I find that either Stata or SPSS can give me the results I want. Overall though, I feel that Stata’s statistical commands are generally far more logical and consistent (and sometimes more powerful) than their SPSS counterparts. --- end of excerpt --- p.s. - Tabular output from Stata is pure text, so anyone looking for pivot tables will no doubt agree with John on his point about output.
--
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/). |
|
|
Another two penn’orth: Bring back: 1: /format condense in FREQUENCIES (see slide 40 in http://surveyresearch.weebly.com/uploads/2/9/9/8/2998485/slides_2_-_layout_usage_and_changes.ppt) 2: /crossbreak in MEANS. (see slides 47 and 48: slide 48 is exactly the basic output needed for elaboration as each cell displays a % and its base n, but the epsilons still have to be calculated manually) 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 |
|
|
How is CROSSBREAK any better than CTABLES? CTABLES /TABLE jobcat [C] > salary [S][MEAN, COUNT] BY minority [C] /SLABELS POSITION=ROW /CATEGORIES VARIABLES=jobcat minority TOTAL=YES. On Sat, Jul 16, 2016 at 10:24 PM, John F Hall <[hidden email]> wrote:
|
|
|
Jon I had to dig deep into the past to find this example (and the data) but in the 1980s: MEANS VARIABLES = SEXISM (0,9) SEX (1,2) ETHNIC (1,2) /CROSSBREAK = SEXISM BY SEX BY ETHNIC /CELLS = MEAN COUNT . . . produced the following table: ETHNIC Mean : Count : White Other Row : Total : 1 : 2 : SEX --------:----------:----------: 1 : 4.63 : 3.43 : 3.98 Boys : 19 : 23 : 42 -:----------:----------: 2 : 1.89 : 1.80 : 1.84 Girls : 19 : 25 : 44 -:----------:----------: Column Total 3.26 2.58 2.88 38 48 86 A modified version of your CTABLES syntax: CTABLES /TABLE v348 [C] > sexism [S][MEAN, COUNT] BY ethnic [C] /SLABELS POSITION=ROW /CATEGORIES VARIABLES=v348 ethnic TOTAL=YES. . . produces this table: (The figures differ slightly as I’m not quite sure how the derived variable ETHNIC was calculated from the 16 categories coded.)
IMHO the CROSSBREAK syntax is easier to understand (and write) and the output simpler (ie less cluttered). All students really need is something like this:
(Source: Fifth form survey, 1981) . . from which they can calculate the differences between means and draw tentative conclusions. The data are from a real survey (by three of my sophomore students) and the sample was very small, but it’s the logic that matters. The same analysis can be used when a criterion value of the dependent variable is recoded to 100 and other values to zero: the figures in the table then represent %% and the figures in brackets are (n = 100%). 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: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Jon Peck How is CROSSBREAK any better than CTABLES? On Sat, Jul 16, 2016 at 10:24 PM, John F Hall <[hidden email]> wrote:
-- Jon K Peck ===================== 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 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Administrator
|
John, if you want a command called CrossBreak giving the type of very simple output you showed, you could always use a macro. E.g., give this a try:
DEFINE CrossBreak ( Y = !CHAREND('/') / RowVar = !CHAREND('/') / ColVar = !CMDEND ). CTABLES /TABLE !RowVar [C] > !Y [S][MEAN, COUNT] BY !ColVar [C] /SLABELS POSITION=ROW VISIBLE=NO /CATEGORIES VARIABLES= !RowVar !ColVar TOTAL=YES /TITLES CAPTION ='NOTE: Cells show MEAN and COUNT' . !ENDDEFINE. * Read in some data to illustrate. * Modify path on the FILE HANDLE command as needed. FILE HANDLE TheDataFile /NAME="C:\SPSSdata\survey_sample.sav". NEW FILE. DATASET CLOSE all. GET FILE = "TheDataFile". * Call the macro. CrossBreak Y = age / RowVar = sex / ColVar = race. And before you object that your students don't know how to write macros, they don't have to know how to write them. They just have to be adept enough to run the macro definition once per session.
--
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/). |
|
|
Bruce I was about to ask the list if a macro would be appropriate, but you beat me to it. Thanks. I'll test it out on some real data: it's exactly what I've been asking for since way back when. Your example is what I call first order in my tutorial 3.2.4 Income differences – Elaboration http://surveyresearch.weebly.com/uploads/2/9/9/8/2998485/3.2.4b__income_differences__elaboration.pdf but it would be nice if it could be extended to second, third or higher order tables. Ultimately I would like to be able to produce tables such as: Model X → Y . T1 T2 Dependent variable Y = Earnings group (<£6000 pa, <£12,000 pa, £12,000+ pa) Independent variable X = Sex (Men, Women) Test variable 1 T1 = Hours of work (Part-time, Full-time Test variable 2 T2 = Type of work (Non-manual, manual) * Earnings from paid work of men and women working full time (30 or more hours a week). temp. select if workmode = 2. ctables /VLABELS VARIABLES=sex incr3 class DISPLAY=NONE /table sex by class > incr3 [ROWPCT.COUNT f8.1 "%" TOTALS[COUNT f8.0 "n= 100%"]] /CATEGORIES VARIABLES= sex class incr3 TOTAL=YES POSITION=AFTER. Earnings from paid work of men and women working full time (30 or more hours a week)
(NB: Epsilons calculated manually) The Rolls Royce macro would be able to produce the following:
(Source: British Social Attitudes, 1986) You’ll probably tell me to use logistic regression modelling instead, but students have to start somewhere. 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 -----Original Message----- John, if you want a command called CrossBreak giving the type of very simple output you showed, you could always use a macro. E.g., give this a try: DEFINE CrossBreak ( Y = !CHAREND('/') / RowVar = !CHAREND('/') / ColVar = !CMDEND ). CTABLES /TABLE !RowVar [C] > !Y [S][MEAN, COUNT] BY !ColVar [C] /SLABELS POSITION=ROW VISIBLE=NO /CATEGORIES VARIABLES= !RowVar !ColVar TOTAL=YES /TITLES CAPTION ='NOTE: Cells show MEAN and COUNT' . !ENDDEFINE. * Read in some data to illustrate. * Modify path on the FILE HANDLE command as needed. FILE HANDLE TheDataFile /NAME="C:\SPSSdata\survey_sample.sav". NEW FILE. DATASET CLOSE all. GET FILE = "TheDataFile". * Call the macro. CrossBreak Y = age / RowVar = sex / ColVar = race. And before you object that your students don't know how to write macros, they don't have to know how to write them. They just have to be adept enough to run the macro definition once per session. John F Hall wrote > Jon > > I had to dig deep into the past to find this example (and the data) > but in the 1980s: > > MEANS VARIABLES = > SEXISM (0,9) > SEX (1,2) > ETHNIC (1,2) > /CROSSBREAK = > SEXISM BY SEX BY ETHNIC > /CELLS = MEAN COUNT . > > . . produced the following table: > > > ETHNIC > Mean : > Count : White Other Row > : Total > : 1 : 2 : > SEX --------:----------:----------: > 1 : 4.63 : 3.43 : 3.98 > Boys : 19 : 23 : 42 > -:----------:----------: > 2 : 1.89 : 1.80 : 1.84 > Girls : 19 : 25 : 44 > -:----------:----------: > Column Total 3.26 2.58 2.88 > 38 48 86 > > A modified version of your CTABLES syntax: > > CTABLES > /TABLE v348 [C] > sexism [S][MEAN, COUNT] BY ethnic [C] > /SLABELS POSITION=ROW > /CATEGORIES VARIABLES=v348 ethnic TOTAL=YES. > > . . produces this table: > > (The figures differ slightly as I’m not quite sure how the derived > variable ETHNIC was calculated from the 16 categories coded.) > > > > ethnic > > 1 White > 2 Other > Total > > v348 Q.39 Sex > 1 Boys > sexism > Mean > 4.43 > 3.41 > 3.85 > > Count > 21 > 27 > 48 > > 2 Girls > sexism > Mean > 1.87 > 1.76 > 1.81 > > Count > 23 > 29 > 52 > > Total > sexism > Mean > 3.09 > 2.55 > 2.79 > > Count > 44 > 56 > 100 > > IMHO the CROSSBREAK syntax is easier to understand (and write) and the > output simpler (ie less cluttered). All students really need is > something like this: > > > Mean sexism > (n) > > ethnic > > White > Other > Total > > > Boys > 4.43 > 3.41 > 3.85 > > (21) > (27) > (48) > > Girls > 1.87 > 1.76 > 1.81 > > (23) > (29) > (52) > > Total > 3.09 > 2.55 > 2.79 > > (44) > (56) > (100) > > (Source: Fifth form survey, 1981) > > . . from which they can calculate the differences between means and > draw tentative conclusions. The data are from a real survey (by three > of my sophomore students) and the sample was very small, but it’s the > logic that matters. The same analysis can be used when a criterion > value of the dependent variable is recoded to 100 and other values to > zero: the figures in the table then represent %% and the figures in brackets are (n = 100%). > > John F Hall (Mr) > [Retired academic survey researcher] > > Email: > johnfhall@ > <mailto: > johnfhall@ > > > Website: www.surveyresearch.weebly.com > <http://www.surveyresearch.weebly.com/> > SPSS start page: > www.surveyresearch.weebly.com/1-survey-analysis-workshop > <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html&g > t; > > > > > > > > From: SPSSX(r) Discussion [mailto: > ] On Behalf Of Jon Peck > Sent: 17 July 2016 15:52 > To: > Subject: Re: Making SPSS better > > How is CROSSBREAK any better than CTABLES? > > > > On Sat, Jul 16, 2016 at 10:24 PM, John F Hall < > johnfhall@ > <mailto: > johnfhall@ > > > wrote: > Another two penn’orth: > > Bring back: > > 1: /format condense in FREQUENCIES > (see slide 40 in > http://surveyresearch.weebly.com/uploads/2/9/9/8/2998485/slides_2_-_la > yout_usage_and_changes.ppt) > > 2: /crossbreak in MEANS. > (see slides 47 and 48: slide 48 is exactly the basic output needed for > elaboration as each cell displays a % and its base n, but the epsilons > still have to be calculated manually) John F Hall (Mr) [Retired > academic survey researcher] > > Email: > johnfhall@ > <mailto: > johnfhall@ > > > Website: www.surveyresearch.weebly.com > <http://www.surveyresearch.weebly.com/> > SPSS start page: > www.surveyresearch.weebly.com/1-survey-analysis-workshop > <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html&g > t; > > ===================== To manage your subscription to SPSSX-L, send a > message to > <mailto: > > (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 > > > > > -- > Jon K Peck > jkpeck@ > <mailto: > jkpeck@ > > > ===================== To manage your subscription to SPSSX-L, send a > message to > <mailto: > > (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 > (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 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/Making-SPSS-better-tp5732642p5732775.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 |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Administrator
|
TRIM YOUR POSTS!!!!!
"The Rolls Royce macro would be able to produce the following:" Blah blah blah.... Become a master of the AGGREGATE command and COMPUTE then you will never have to manually calculate anything. WTF is this Epsilon you find so important? Maybe provide a formula and a rationale and perhaps one of the master SPSS proggers will provide the "The Rolls Royce macro" If you can do it manually then a few lines of code will do it too.
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?" |
|
|
Q: WTF is this Epsilon you find so important? (DM)
"Elaboration relies on epsilon (percentage point difference) and is best used with dichotomised data, but can be used to compare any two categories of variables with three or more categories. It’s not particularly sophisticated as it loses information when categories are condensed, but it was good enough for Rosenberg. It’s easily understood by beginners, simple to specify tables in SPSS CROSSTABS and very useful for making students think about effects and interactions. CTABLES is perhaps too complex to specify for beginners, but the tables are much more useful."
(See page 14 of 3.2.4 Income differences – Elaboration http://surveyresearch.weebly.com/uploads/2/9/9/8/2998485/3.2.4b__income_differences__elaboration.pdf)
If you had bothered to read my tutorial (RTFT?) you would have found detailed work-throughs for the following research questions:
1: Is there a difference between the earnings (from paid work) of men and women?
2: What other variables might account for differences in earnings?
3: What effect do they have by themselves?
4: What happens to any differences in earnings between men and women when controlling for these other variables?
I fail to see how AGGREGATE can calculate the points difference between two percentages in the same row or column of a table.
|
|
|
concentrate on the simple things as well as the trendy new bits, BUT on trendy new bits, for someone who uses logic a lot
Would like to see a list of requests from this list from SPSS to see which, if any of the requests get acted on
best
Diana
Work
University of
Hertfordshire
College Lane,
Hatfield, Hertfordshire AL10 9AB, UK
+44 (0) 170 728 4626
skype:
kornbrotme
Home
19 Elmhurst Avenue
London N2 0LT,
UK
+44
(0) 208 444 2081
===================== 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 |
| Free forum by Nabble | Edit this page |